eu support to timber‑producing countries under the flegt action … · 2019-09-22 · executive...
TRANSCRIPT
Special Report The certification bodies’ new role on CAP expenditure: a positive step towards a single audit model but with significant weaknesses to be addressed
(pursuant to Article 287(4), second subparagraph, TFEU)
EN 2017 NO 07
1977 - 2017
The ECA’s special reports set out the results of its performance and compliance audits of specific budgetary areas or management topics. The ECA selects and designs these audit tasks to be of maximum impact by considering the risks to performance or compliance, the level of income or spending involved, forthcoming developments and political and public interest.
Audit team
This report was adopted by Audit Chamber I of the ECA — headed by ECA Member Phil Wynn Owen — which is responsible for the audit for sustainable use of Natural Resources. The audit was led by ECA Member João Figueiredo, supported by Paula Betencourt, private office attaché; Sylvain Lehnhard, acting principal manager; Luis Rosa, head of task and Marius Cerchez, deputy head of task. The audit team consisted of Ioan Alexandru Ilie, Jindrich Dolezal, Michal Machowski, Michael Spang, Antonella Stasia, Heike Walz. Michael Pyper assisted with drafting the report.
From left to right: Michael Pyper, Paula Betencourt, Sylvain Lehnhard, João Figueiredo, Luis Rosa, Heike Walz, Jindrich Dolezal.
2
CONTENTS
Paragraph
Glossary
Abbreviations
Executive Summary I - VII
Introduction 1 - 12
Expenditure under the Common Agricultural Policy 1 - 3
The legal and institutional framework for shared management under the CAP 4 - 7
Role and responsibilities of certification bodies 8 - 12
Audit scope and approach 13 - 21
Audit scope 13 - 17
Audit approach 18 - 21
Observations 22 - 85
The CBs’ new role in relation to the legality and regularity of CAP expenditure is a positive step towards a single audit model 22 - 28
DG AGRI’s assurance model remains based on the Member States’ control results, 29- 37
Risk assessment and sampling method 38 - 61
Basing the risk assessment on the accreditation matrix may inflate the level of assurance the CBs can derive from PAs’ internal control systems 38 - 47
The CBs’ sampling method for IACS transactions, based on the PAs’ lists of randomly selected on-the-spot checks, entailed a series of risks that were not overcome 48 - 58
A portion of the non-IACS transactions upon which the CBs perform substantive testing is not representative of expenditure in the financial year audited 59 - 61
Substantive testing 62 - 71
3
For most transactions, the guidelines do not require the CBs to perform on-the-spot testing at final beneficiary level 62 - 67
The guidelines require the CBs only to re-perform PAs’ initial checks, rather than carry out all the audit procedures the CBs consider necessary to obtain reasonable assurance 68 - 71
Conclusion and audit opinion 72 - 85
The guidelines require the CBs to calculate two different error rates and the use made of these rates by the CBs and DG AGRI is not appropriate 72 - 78
The CBs’ opinion on the legality and regularity of expenditure is based on an understated total error 79 - 85
Conclusions and recommendations 86 - 104
Annex I - The Commission’s shared management model, as presented in its 2015 AAR
Annex II - The Commission’s assurance model for legality and regularity of expenditure, as presented in its 2015 AAR
Annex III - An example of an accreditation matrix, adapted from DG AGRI’s Guideline No 3 for the certification audit of the accounts
Annex IV - The timetable of the PAs’ and CBs’ substantive testing and reporting to the Commission
The Commission’s replies
4
GLOSSARY
Accreditation: A process to certify that paying agencies have an administrative organisation
and internal control system which provide sufficient guarantees that payments are legal and
regular, and properly accounted for. Such certification is to be made by MSs on the basis of
the paying agencies’ compliance with a set of criteria (“accreditation criteria”) regarding
internal environment, control activities, information and communication, monitoring.
Adjusted error rate: This is the Commission’s estimate of the level of residual error affecting
the CAP payments by PAs to beneficiaries after all checks have been carried out. The
Commission calculates them and publishes them in its AAR.
Administrative checks: Formalised documentary checks carried out by PAs on all
applications in order to verify that they comply with the terms under which aid is granted.
For IACS expenditure, the information contained in IT databases is used for automatic cross-
checks.
Annual Activity Report (AAR): An annual report published by every Directorate-General
detailing its achievements, initiatives it has taken during the year and the resources it has
used. DG AGRI’s annual activity report also includes an assessment of the functioning and
the results of the CAP management and control systems at PA level.
Conformity clearance procedure: Commission’s procedure based on an assessment of the
PAs’ internal control systems, to ensure that Member States apply EU and national law and
that any expenditure that infringes these rules in one or more financial years is excluded
from EU financing via a financial correction.
Control statistics: Annual reports submitted by Member States to the Commission that
contain the results of the paying agencies’ administrative and on-the-spot checks.
European Agricultural Fund for Rural Development (EAFRD): Finances the Union’s financial
contribution to rural development programmes.
5
European Agricultural Guarantee Fund (EAGF): Provides funding for direct payments to
farmers, for the management of agricultural markets and for information and promotion
measures.
Financial clearance procedure: Results in Commission’s annual financial decision concerning
the completeness, accuracy and veracity of the annual accounts of each accredited PA.
Financial corrections: Commission’s exclusions from EU financing of expenditure that
Member States have not effected in conformity with the applicable EU and national law. For
CAP expenditure, the exclusions always take the form of financial corrections treated as
assigned revenue.
Integrated Administration and Control System (IACS): An integrated system consisting of
databases of holdings, applications, agricultural areas, animals and, where applicable,
payment entitlements. These databases are used for administrative cross-checks on aid
applications for area and animal related payments.
Internationally accepted audit standards (IAAS) comprise auditing standards specified by
different public and professional standard-setting bodies, such as the International
Standards on Auditing (ISAs) issued by the International Auditing and Assurance Standards
Board (IAASB) or the International Standards of Supreme Audit Institutions (ISSAI) issued by
the International Organisation of Supreme Audit Institutions (INTOSAI).
Management declaration (MD): A yearly declaration of the Directors of each PA as to the
completeness, accuracy and veracity of the accounts and the proper functioning of the
internal control systems, as well as to the legality and regularity of the underlying
transactions. Several annexes accompany the MD, among which one regarding the statistics
of the results of all the administrative and on-the-spot checks performed by the PA (“the
control statistics”).
On-the-spot checks: Checks carried out by PAs’ inspectors in order to verify that the final
beneficiaries comply with the applicable rules. These can take the form of classical on-the-
spot visits (to agricultural holdings or investments subject to aid) or of remote sensing
6
(review of recent satellite images of parcels) to be complemented with rapid field visits in
cases of doubts.
Paying agency (PA): The body(ies) responsible within a Member State for the management
and control of CAP expenditure, notably controls, calculation and payment of CAP aid to the
beneficiaries and their reporting to the Commission. Part of a PA’s work may be done by
delegated bodies, but not payments to beneficiaries and their reporting to the Commission.
Programme period: Multiannual framework for planning and implementing EU policies,. The
current programme period runs from 2014 to 2020 Rural development programmes funded
by EAFRD are managed under this multiannual framework; EAGF is managed on a yearly
basis.
7
ABBREVIATIONS
AAR Annual activity report
CA Competent Authority of the Member State (usually the Ministry of
Agriculture)
CAP Common Agricultural Policy
CB Certification body
CF Cohesion Fund
DG AGRI Directorate General for Agriculture and Rural Development
EAFRD European Agricultural Fund for Rural Development
EAGF European Agricultural Guarantee Fund
ERDF European Regional Development Fund
ERR Error rate
ESF European Social Fund
IACS Integrated Administrative and Control System
IRR ‘Incompliance’ rate
ISAs International Standards on Auditing
ISSAIs International Standards of Supreme Audit Institutions
MD Management Declaration
PA Paying Agency
8
EXECUTIVE SUMMARY
I. With a budget of 363 billion euro (in 2011 prices) for the 2014-2020 period
(around 38 % of the total amount of the multiannual financial framework 2014-2020), the
Common Agricultural Policy (CAP) is under shared management between the Commission
and the Member States (MSs). While the Commission delegates the implementation of the
budget to paying agencies (PAs) designated by MSs, it retains ultimate responsibility and is
required to ensure that the EU’s financial interests are protected as though the Commission
were performing the delegated budget implementation tasks itself.
II. For this purpose, certification bodies (CBs) appointed by MSs have been entrusted with
the role of being the PAs’ independent auditors since 1996. They have been required, since
the 2015 financial year, to also provide an opinion, prepared in accordance with
internationally accepted audit standards, stating whether the expenditure for which
reimbursement has been requested from the Commission is legal and regular. The 2015
financial year was thus the first for which the Commission could use the CBs’ enhanced work
on legality and regularity.
III. In this context, we assessed whether the framework set up by the Commission enabled
the CBs to draw an opinion on the legality and regularity of CAP expenditure in accordance
with the applicable EU regulations and internationally accepted audit standards, thus
providing reliable results for Commission’s assurance model. We conclude that the
framework designed by the Commission for the first year of implementation of the new
work of the CBs has significant weaknesses. As a result, the CBs’ opinions do not fully
comply with the applicable standards and rules in important areas.
IV. The CBs' new role is a positive step towards a single audit model because the CBs'
outputs have the potential to help MSs to strengthen their control systems, reduce audit and
control costs and enable the Commission to obtain independent additional assurance as to
the legality and regularity of expenditure.
V. Against this backdrop, we analysed the Commission’s existing assurance model and the
changes introduced to take account of the CBs’ enhanced role. We noted that the
Commission’s assurance model remains based on the Member States’ control results. For
9
the 2015 financial year, the CBs’ opinion on legality and regularity were merely one factor
taken into account when the Commission calculated its adjustments of the errors reported
in the Member States control statistics. As those opinions are the only source which provides
independent assurance on legality and regularity on an annual basis, the CBs’ work, once
done in a reliable manner, should become the key element for the Commission’s assurance.
VI. Our examination of the Commission’s guidelines compliance with the applicable
regulations and internationally accepted audit standards identified the following
weaknesses:
(a) For the risk assessment procedure, we observed that the Commission required CBs to
use the accreditation matrix which creates the risk of inflating the level of assurance
that the CBs can derive from PAs’ internal control systems;
(b) The CBs’ sampling method for transactions based on the PAs’ lists of randomly selected
on-the-spot checks entailed a series of additional risks that were not overcome: in
particular, the CBs’ work can only be representative if the samples initially selected by
the PAs are themselves representative. A portion of the sampling for non-IACS
transactions is not representative of expenditure and is thus is not representative of the
financial year audited;
(c) As regards substantive testing, the Commission required CBs, for part of the sampling,
to carry out only a re-performance of the PAs' administrative checks;
(d) The Commission required CBs for their substantive testing, only to re-perform the PAs’
initial checks. While re-performance is a valid audit collection method, internationally
accepted audit standards also state that auditors should choose and perform all audit
steps and procedures that they themselves consider appropriate in order to obtain
sufficient audit evidence to form a reasonable assurance opinion;
(e) As regards the drawing of the auditor’s conclusion, we found that the guidelines
required the CBs to calculate two different error rates in relation to legality and
regularity, and that the use made of these rates not only by the CBs, but also by DG
AGRI, was not appropriate;
10
(f) Finally, we concluded that the CBs’ opinion on the legality and regularity of expenditure
was based on an understated total error. Indeed, the errors detected and reported by
the PAs in their control statistics were not taken into account by the CBs in calculating
their estimated level of error.
VII. We make a number of recommendations to address these observations:
(a) The Commission should use the CBs’ results, when the work is defined and performed in
accordance with the applicable regulations and internationally accepted audit
standards, as the key elementof its assurance model regarding the legality and
regularity of expenditure;
(a) The Commission should revise its guidelines as follows:
(i) focus the CBs’ risk assessment as regards legality and regularity on the key and
ancillary controls already used by the Commission;
(ii) require CBs, for the selection by the CBs of IACS transactions from the list of claims
randomly selected by the PAs for on-the-spot checks, to put in place appropriate
safeguards concerning the representativeness of the CBs’ samples; the timely CBs
on-the-spot visits; the non-disclosure of the CBs’ sample to the PAs;
(iii) for the sampling of non-IACS expenditure, require the CBs to draw their samples
directly from the lists of payments;
(iv) allow the CBs to carry out: on-the-spot testing for any transaction audited, and to
carry out all audit steps and procedures that they themselves consider appropriate,
without being limited to re-performing the PAs initial checks;
(v) require the CBs to calculate only a single error rate regarding legality and
regularity;
(vi) for IACS transactions which are sampled from the list of PAs’ random on-the-spot
checks, the overall error calculated by the CBs, to be able to issue an opinion on
legality and regularity of expenditure, should also include the level of error
reported by the PAs in the control statistics, extrapolated to the remaining
11
transactions not subject to PA on-the-spot checks . The CBs have to ensure that the
control statistics compiled by the PAs are complete and accurate.
12
INTRODUCTION
Expenditure under the Common Agricultural Policy
For agriculture, EU support is granted through the European Agricultural Guarantee 1.
Fund (EAGF) and the European Agricultural Fund for Rural Development (EAFRD). The total
allocation for these two funds amounts to 363 billion euro (in 2011 prices) for the 2014-2020
programme period, which represents around 38 % of the total MFF for the period 2014-
20201
Common Agricultural Policy (CAP) expenditure under both funds can be grouped in two 2.
categories:
.
- Integrated Administrative and Control System (IACS) expenditure, which is entitlement-
based2
- non-IACS expenditure, which are reimbursement-based payments, consisting mainly of
investments in farms and rural infrastructure, as well as interventions in agricultural
markets.
and consists mainly of annual per hectare payments.;
The split between IACS and non-IACS expenditure for the two funds (around 86 % and 3.
14 % of CAP expenditure respectively) for the 2015 financial year is show in Figure 1
1 Annex I to Council Regulation (EU, Euratom) No 1311/2013 of 2 December 2013 laying down the multiannual financial framework for the years 2014-2020 (OJ L 347, 20.12.2013, p. 884).
below:
2 Where a payment is based on meeting certain conditions.
13
Figure 1 – The breakdown of CAP expenditure for the 2015 financial year (in billion euro)
The legal and institutional framework for shared management under the CAP
Responsibility for managing the CAP is shared between the Commission and the 4.
Member States. About 99 % of the CAP budget is spent under shared management, as
defined in the Financial Regulation3. The CAP Horizontal Regulation4 lays down ground rules
specifically for the financing, management and monitoring of the CAP. The Commission has
been empowered to further specify these rules through implementing regulations5
3 Article 59 of Regulation (EU, Euratom) No 966/2012 of the European Parliament and of the Council of 25 October 2012 on the financial rules applicable to the general budget of the Union and repealing Council Regulation (EC, Euratom) No 1605/2002 (OJ L 298, 26.10.2012, p. 1) (the “Financial Regulation”).
and
guidelines.
4 Regulation (EU) No 1306/2013 of the European Parliament and of the Council of 17 December 2013 on the financing, management and monitoring of the common agricultural policy and repealing Council Regulations (EEC) No 352/78, (EC) No 165/94, (EC) No 2799/98, (EC) No 814/2000, (EC) No 1290/2005 and (EC) No 485/2008 (OJ L 347, 20.12.2013, p. 549) (the “CAP Horizontal Regulation”).
5 Commission Implementing Regulation (EU) No 809/2014 of 17 July 2014 laying down rules for the application of Regulation (EU) No 1306/2013 of the European Parliament and of the Council with regard to the integrated administration and control system, rural development measures
42.2
6.3
2.7
5.4
0 10 20 30 40 50
EAGF
EAFRD
IACS
non-IACS
14
Under shared management, the Commission remains ultimately responsible for the 5.
budget, but delegates its implementation to bodies specially designated by the Member
States: the paying agencies (PAs). The Commission’s Directorate-General for Agriculture and
Rural Development (DG AGRI) oversees the PAs’ implementation of the budget, verifying and
reimbursing the expenditure declared monthly (for the EAGF) and quarterly (for the EAFRD)
by the PAs and, ultimately, assessing whether expenditure is legal and regular by means of
the conformity clearance procedure6. If this procedure reveals weaknesses in a PA’s control
system which have a financial impact on the EU budget, the Commission may apply financial
corrections to that PA. DG AGRI’s shared management model, setting out the roles of the
different parties involved, is presented in Annex I7
At Member State level, the CAP budget is managed by PAs accredited by the relevant 6.
Member State’s competent authority (CA)
.
8. The PAs perform administrative checks on all
project applications and payment claims received from beneficiaries, as well as on-the-spot
checks on a minimum sample of 5 % for the vast majority of supporting measures9
and cross compliance (OJ L 227, 31.7.2014, p. 69) and Commission Implementing Regulation (EU) No 908/2014 of 6 August 2014 laying down rules for the application of Regulation (EU) No 1306/2013 of the European Parliament and of the Council with regard to paying agencies and other bodies, financial management, clearance of accounts, rules on checks, securities and transparency (OJ L 255, 28.8.2014, p. 59).
.
Following these checks, the PAs pay the beneficiaries the amounts due and declare those
amounts monthly (EAGF)/quarterly (EAFRD) to the Commission for reimbursement. All the
amounts paid are then recorded in the PAs’ annual accounts. The director of each PA
provides the Commission with these annual accounts, along with their declaration (the
6 DG AGRI procedure based on an assessment of the PAs’ internal control systems, to ensure that Member States apply EU and national law and that any expenditure that infringes these rules in one or more financial years is excluded from EU financing (Article 52 of the CAP Horizontal Regulation).
7 CAP shared management model presented in DG AGRI 2015 AAR, p. 35.
8 As required by Article 7 of the CAP Horizontal Regulation.
9 Articles 30 to 33 of Regulation (EU) No 809/2014. For the 2014-2020 period, the Commission has introduced further possibilities for Member States to reduce the minimum control rate down to 1 %. Member States may use such possibilities as provided by Article 41 of the Regulation (EU) No 908/2014 and by Article 36 of the Regulation (EU) No 809/2014.
15
management declaration (MD)) regarding the effectiveness of their control systems, which
also summarises the levels of error stemming from their control statistics10
The Financial Regulation requires the Commission, in entrusting the PAs with budget 7.
implementation tasks for the CAP under shared management, to ensure that the EU’s
financial interests are protected to the same standard as though the Commission were
performing these tasks itself
. The
Commission’s authorising officer (DG AGRI’s Director General for the CAP) takes these
annual accounts and MDs into account in DG AGRI’s annual activity report (AAR).
11
Role and responsibilities of certification bodies
. This includes establishing appropriate control and audit
responsibilities, such as the examination and acceptance of accounts. For the CAP, the
certification body (CB) assumes these audit responsibilities at Member State level.
The CBs have been performing their role as independent auditors of the PAs since the 8.
1996 financial year12
The Financial Regulation
. The role required the CBs to issue a certificate, in accordance with
internationally accepted audit standards, stating whether the accounts to be transmitted to
the Commission were true, complete and accurate, and whether the internal control
procedures had operated satisfactorily, with particular reference to the accreditation
criteria.
13 and the CAP Horizontal Regulation14
10 Annual reports submitted by Member States to the Commission that contain the results of the paying agencies’ administrative and on-the-spot checks.
have increased the role 9.
and responsibilities of CBs. In addition to their responsibility for the accounts and internal
control, CBs have been required, since the 2015 financial year, to provide an opinion stating
11 Article 59 and recital 22 of the Financial Regulation.
12 Article 3 of Commission Regulation (EC) No 1663/95 of 7 July 1995 laying down detailed rules for the application of Council Regulation (EEC) No 729/70 regarding the procedure for the clearance of accounts of the EAGGF Guarantee Section (OJ L 158, 8.7.1995, p. 6).
13 Article 59(5) of the Financial Regulation.
14 Article 9 of the CAP Horizontal Regulation.
16
whether the expenditure for which reimbursement has been requested from the
Commission is legal and regular. This opinion, drawn up in accordance with internationally
accepted audit standards, must also state ‘whether the examination puts in doubt the
assertions made in the MD’.
The CAP horizontal regulation also stipulates that a CB must be a public or private audit 10.
body designated by the CA and that it must have the necessary technical expertise and be
operationally independent from the PA, as well as from the authority which has accredited
that PA. For the 2015 financial year, CAP expenditure in the 28 Member States was
administered by a total of 80 PAs, which were in turn audited by 64 CBs15. Forty six of these
CBs were public bodies16
The aim of the CBs’ work on legality and regularity is to increase DG AGRI’s assurance on 11.
the legality and regularity of expenditure. The CBs’ role in DG AGRI’s assurance model is
described in
and 18 were private audit companies.
Annex II
The Commission established the framework for the CBs’ audit work through an 12.
implementing regulation
. Possible uses of the CBs’ work in the context of a “single audit model”
are detailed in paragraphs 24 to 28.
17 and additional guidelines18
15 In several Member States, more than one paying agency is audited by the same certification body.
. According to this regulation, and in
order to obtain reliable reports and opinions from the CBs, the Commission needs to ensure
that:
16 Such as departments within Ministries of Finance, departments of regional governments, or Supreme Audit Institutions.
17 Articles 5 to 7 of the Regulation (EU) No 908/2014.
18 Guideline no. 1 on accreditation requirements; Guideline no. 2 on the annual certification audit; Guideline no. 3 on reporting requirements; Guideline no. 4 on the Management Declaration; and Guideline no. 5 on irregularities.
17
(a) its guidelines provide appropriate instructions in accordance with internationally
accepted audit standards19
(b) the work carried out by the CBs based on such guidance is sufficient and of appropriate
quality.
;
AUDIT SCOPE AND APPROACH
Audit scope
The new requirement for CBs to provide an opinion on the legality and regularity of the 13.
expenditure plays a key role in the Commission’s assurance model for the 2014-2020 period.
In this audit, we focused our analysis on the framework established by the Commission 14.
for the CBs’ work on legality and regularity for the first year of implementation (2015
financial year). This audit did not assess the substantive testing carried out by the CBs.
The audit aimed at assessing whether the CBs’ new role was a step towards a single 15.
audit approach and if the Commission took due account of it in its assurance model. The
audit also aimed at assessing whether the framework set up by the Commission enables the
CBs to draw an opinion on the legality and regularity of CAP expenditure in accordance with
the applicable EU regulations and internationally accepted audit standards.
In particular, we examined whether the Commission’s guidelines to CBs ensured : 16.
- an appropriate risk assessment by the CBs and a representative sampling;
- an appropriate level of substantive testing; and
- a correct estimate of the level of error and audit opinion.
After the finalisation of our audit field work (September 2016), the Commission finalised 17.
in January 2017 new guidelines to be applied by CBs from 2018 financial year. These revised
guidelines have not been subject to our review.
19 Article 9 of the CAP Horizontal Regulation.
18
Audit approach
The audit evidence was obtained at both Commission (DG AGRI) and Member State 18.
level. The starting point for the work was an analysis of the relevant legal framework, in
order to identify the requirements applicable to all parties involved: the Commission, the
CAs and the CBs.
At Commission level, the audit involved: 19.
- a review of the Commission’s guidelines for CBs applicable to the 2015 financial year, as
well as documents from supporting activities such as expert group meetings20
- an assessment of the Commission’s desk reviews of CBs’ reports and opinions for 25 PAs
in 17 Member States, which in 17 cases also included a review of DG AGRI’s reports on
audit visits to CBs. We also reviewed DG AGRI’s 2015 AAR and the use made therein of
CBs’ work on legality and regularity;
(we
participated as observers in one such meeting in Brussels on 14 and 15 June 2016);
- interviews with relevant DG AGRI staff .
At Member State level, the audit involved: 20.
- a survey of 20 CAs and CBs in 13 Member States21
20 Expert group meetings were organised by DG AGRI, with the involvement of representatives from CAs and CBs from all Member States, as a forum for sharing knowledge and experience and clarifying issues related to the CBs’ work.
, selected based on the value of the
expenditure they certified (in total, the selection accounted for 63 % of EU CAP
expenditure in 2014) and on the CB’s legal status (i.e. whether it was a public or private
entity). The topics covered by the surveys were amongst others the Commission’s
guidance and support to the CBs;
21 Belgium, Czech Republic, Germany, Ireland, Spain, France, Italy, Austria, Poland, Portugal, Romania, Slovakia and United Kingdom.
19
- six visits22
Our findings were examined against the relevant provisions in the regulations for the 21.
2014-2020 programme period and the requirements laid down in the relevant International
Standards of Supreme Audit Institutions (ISSAIs) or International Standards on Auditing
(ISAs).
to Member States, carrying out interviews with representatives from the CAs
and CBs for the purpose of better understanding the information provided in the replies
to the survey, as well as examining the CBs’ contractual arrangements, checking the
timeliness of their audit procedures and reviewing the main results of their work as
presented in their audit reports and opinions covering the 2015 financial year.
OBSERVATIONS
The CBs’ new role in relation to the legality and regularity of CAP expenditure is a positive
step towards a single audit model
While there is no single recognised definition of the term “single audit”, the concept is 22.
premised upon the need to avoid uncoordinated, overlapping controls and audits. In the
context of the EU budget, and for shared management in particular, “single audit” would
provide assurance as to the legality and regularity of expenditure on two fronts: for
management (internal control) purposes on the one hand and for audit purposes on the
other. A single audit approach should be an effective control framework in which each layer
builds on the assurance provided by others.
In the area of Cohesion23, Member State audit authorities have been, since the 2007-23.
2013 period, one of the main sources of information in the Commission’s assurance
framework for the legality and regularity of expenditure. Also in our Opinion No 2/200424
22 Germany (Bavaria), Spain (Castilla y Léon), Italy (AGEA), Poland, Romania and United Kingdom (England).
,
we analysed the “single audit” model for the purposes of internal control, i.e. as a source of
23 Cohesion policy is funded through the European Regional Development Fund (ERDF), the European Social Fund (ESF) and the Cohesion Fund (CF).
24 OJ C 107, 30.4.2004, p. 1.
20
assurance for the Commission’s management. Furthermore, in our Special Report
No 16/201325
For CAP expenditure, the Commission draws assurance that expenditure is legal and 24.
regular from three layers of information: the checks performed by the PAs, the audit work
carried out by the CBs and the results of the Commission’s checks thereon. When such
assurance is considered insufficient, the Commission exercises its ultimate responsibility for
legality and regularity through its own conformity clearance procedures which may result in
financial corrections.
, we made specific recommendations on improving the Commission’s
procedures for using the work of national audit authorities in the area of Cohesion.
The CBs’ new role of issuing an opinion on the legality and regularity of expenditure has 25.
the potential, during the 2014-2020 period, to considerably strengthen a “single audit”
model for the Commission’s management of agricultural expenditure. According to DG AGRI,
such a model will yield greater assurance that control systems are working well26 and that it
will make it possible to derive additional management assurance from the audit work carried
out by the CBs rather than from DG AGRI’s own checks27
The CBs’ audit work can enable a “single audit” model to operate efficiently for CAP 26.
expenditure:
alone.
- by helping Member States to strengthen their control systems;
- by helping to reduce audit and control costs;
- by enabling the Commission to obtain independent additional assurance as to the
legality and regularity of expenditure.
25 http://eca.europa.eu
26 In DG AGRI’s audit manual version of 4.11.2015 - section 4.4 – conformity audit on legality and regularity of expenditure for EAGF and EAFRD.
27 In DG AGRI Directorate J’s Multi-annual Working Programme 2015-2018.
21
We consider a “single audit” approach is the right way to obtain assurance as to the 27.
legality and regularity of expenditure at Member State level. Under EU law, just as the
European Court of Auditors acts as the external auditor for the EU budget as a whole, the
CBs act as operationally independent auditors of agricultural spending at Member
State/regional level. In chronological terms, they are the first auditors to provide assurance
as to the legality and regularity of agricultural spending, and they are the only source of such
assurance at national/regional28 level. CBs draw their assurance from a combination of their
own substantive testing and the PAs’ control systems. In the context of a “single audit”
model for the 2014-2020 CAP, such assurance could potentially be used by the Commission,
as indicated in Figure 2
Figure 2 – A single audit model for the CAP 2014-2020
below, provided an appropriate audit trail is ensured.
Source: European Court of Auditors.
28 Belgium, Germany, Spain, Italy and United Kingdom have decentralised the management of CAP funds through regionalised PAs.
European Commission
(EC)
Paying Agency (PA)
Certification Body (CB)
Final Beneficiaries
Syst
ems
chec
ks
Man
agem
ent
Decl
arat
ion
Administrative and on-the-spot checks
CB report and opinion
Systems audit
Substantive testing
EC review CB work
Management checks
Management/audit report Audit
Management declaration
22
We believe that the Commission, being ultimately responsible for the management of 28.
CAP funds and for the regulatory requirements in force, could make the following use of the
CBs’ work on legality and regularity within the framework of a “single audit” approach, when
this work is reliable:
- improving how it estimates residual error and establishes reservations for its annual
activity report29
- determining whether or not the estimated level of error exceeds 50 000 euro or 2 % of
the relevant expenditure
;
30
- determining more accurately and completely the amounts to be excluded from the
EU budget, making more extensive use of extrapolated financial corrections
and, hence, whether or not to start a conformity clearance
procedure;
31
- assessing whether to ask MSs to review the accreditation status of PAs in cases where
there are insufficient guarantees that payments are legal and regular
;
32
.
DG AGRI’s assurance model remains based on the Member States’ control results
The Financial Regulation33
The assurance model used for this purpose by DG AGRI up until the 2015 financial year, 30.
which determined the amounts at risk from a legality and regularity perspective, took as its
requires DG AGRI’s Director-General to prepare an AAR, 29.
declaring that he has reasonable assurance that the control procedures in place give the
necessary guarantees concerning the legality and regularity of the underlying transactions.
29 Article 66(9) of the Financial Regulation.
30 Articles 34 and 35 of Commission Implementing Regulation (EU) No 908/2014.
31 Article 52 of the CAP Horizontal Regulation.
32 Article 7(2) and (5) of the CAP Horizontal Regulation.
33 Article 66 of Regulation (EU, Euratom) No 966/2012 (the Financial Regulation).
23
starting point the control statistics reported by Member States. Then top-ups were added on
the basis of known weaknesses highlighted by DG AGRI’s conformity audits and also by our
audit work. DG AGRI needed to increase the level of error where it found that part of the
errors were not detected by the PAs and therefore not reflected in MSs’ control results. This
calculation produced an adjusted error rate for each PA and each fund.
The aggregated adjusted error rates for the 2014 and 2015 financial years show that DG 31.
AGRI continued to make adjustments on top of the levels of error reported in the PAs’
controls statistics (see Table 1
Table 1 – Results of adjustments made by DG AGRI on top of the control statistics reported
by the PAs for the 2014 and 2015 financial years
).The information from the CB’s new opinions on legality and
regularity had only a marginal effect on these adjustments for the 2015 financial year:
Fund Financial year Average level of error reported in the PAs’ control
statistics
Aggregated adjusted error rate calculated by DG
AGRI EAGF 2014 0.55 % 2.61 %
2015 0.68 % 1.47 % EAFRD 2014 1.52 % 5.09 %
2015 1.78 % 4.99 %
In its 2015 AAR34
In fact, DG AGRI’s assurance model remains based on the control results. For the 2015 33.
financial year, the CBs' opinions on legality and regularity were merely one factor taken into
DG AGRI explained that it had made “very limited” use of the CBs’ 32.
opinions on legality and regularity. This was mainly due to this being the first year in which
the CBs had produced such opinions, the timing of their work, their lack of technical skills
and legal expertise, their inadequate audit strategies and the insufficient sizes of the
samples audited. In addition to the reasons indicated by DG AGRI, we consider that, based
on our findings, the way the Commission defined the CBs’ work in its guidelines has also an
important impact on the reliability of their work.
34 For the calculation of the 2015 adjusted error rates, DG AGRI used the CBs’ results to top up the PAs error rates in only three cases for the EAGF and nine cases for the EAFRD.
24
account when the Commission calculated its adjustments of the errors reported in the MSs'
control statistics. The CBs’ opinions on legality and regularity are the only source which
provides an independent assurance on legality and regularity on an annual basis. Thus, once
the CBs’ work is done in a reliable manner this independent assurance should in our view
become the key element for the Commission’s Director-General for DG AGRI when assessing
whether expenditure is legal and regular.
The continuing focus on the control statistics stemming from the PAs management and 34.
control systems is confirmed by DG AGRI in its 2015 AAR which stated that “the opinion on
legality and regularity should confirm the level of errors in the management and control
system of the CAP that is operated in the Member States”35. DG AGRI stated further that
“through testing of transactions (based on a statistical sample), the Certification Body
auditors are requested to confirm the level of errors found in the initial eligibility checks
performed by the Paying Agency and, if not confirmed, to give a qualified opinion”36
Some important aspects of the Commission’s guidelines, such as those listed below that 35.
we assess in this report, have thus been specifically put in place to accommodate DG AGRI’s
existing assurance model based on the validation of the PAs’ control results:
.
- use of two samples, with Sample 1 drawn from the list of on-the-spot checks performed
by the PAs (see paragraphs 48 to 58);
- limitation of the CBs’ on-the-spot testing to Sample 1 transactions (see paragraphs 62 to
67);
- limitation of the CBs’ scope to re-perform the PAs’ on-the-spot and administrative
checks (see paragraphs 68 to 71);
- calculating two error rates and DG AGRI’s use of the ‘incompliance’ rate (IRR) instead of
the error rate (ERR) (see paragraphs 72 to 78);
35 Page 66 of the annexes to the DG AGRI 2015 AAR.
36 Page 67 of the annexes to the DG AGRI 2015 AAR.
25
- understatement of the error rates based on which CBs express their opinions (see
paragraphs 79 to 85).
Furthermore, in order to apply the Commission’s model, CBs need to perform a higher 36.
volume of work than what is strictly required by the CAP Horizontal Regulation. This is
because the control statistics summarised in the PAs’ management declarations are
calculated separately for the IACS and non-IACS strata (see paragraph 2). DG AGRI also
calculates the adjusted error rates for each of the two strata. This means that CBs are also
required to validate control statistics at strata level (IACS / non-IACS) and not only at fund
level (EAGF/EAFRD). In order to have sufficient evidence to validate these statistics at strata
level, CBs need to significantly increase the size of their samples compared to a scenario
where only a validation at fund level is required.
The next sections also examine the extent to which the Commission’s guidelines comply 37.
with internationally accepted audit standards at the different stages of the audit process
(risk assessment and sampling method, substantive testing, conclusion and audit opinion).
Risk assessment and sampling method
Basing the risk assessment on the accreditation matrix may inflate the level of assurance
the CBs can derive from PAs’ internal control systems
As per DG AGRI’s guidelines37, the CBs obtain their overall level of assurance on legality 38.
and regularity from two sources: their assessment of the PAs’ control environment and their
substantive testing of transactions. Such a model is in line with internationally accepted
audit standards, which prescribe that the auditor should ‘understand the control
environment and the relevant internal controls and consider whether they are likely to
ensure compliance’38
37 Section 5 of Guideline No 2 on sampling methodology.
.
38 Paragraph 53 of the International Standards of Supreme Audit Institutions (ISSAI) 400 – Fundamental principles of compliance auditing.
26
The higher the CBs rate the PAs’ internal control systems, and hence the greater the 39.
assurance they derive from these systems, the fewer transactions will undergo substantive
testing. Table 2 illustrates this inverse relationship using information from DG AGRI’s
guidelines39
Table 2 – Internal control system rating vs size of samples for substantive testing
:
Internal control system rating
Works well Works Works partially Does not work
Number of items included in substantive testing
93 111 137 181
DG AGRI required the CBs to base their rating of the PAs’ internal control systems on a 40.
matrix for assessing the PAs’ compliance with the accreditation criteria (the “accreditation
matrix”). Only accredited PAs (i.e. those fulfilling the accreditation criteria) may manage CAP
expenditure40
As
.
Annex III41
This overall score is used both by the CAs (to assess whether PAs should keep their 42.
accreditation) and by the CBs (to help them form their opinion as to whether the PAs’
internal control systems are functioning properly).
shows, the accreditation matrix consists of six functions and eight 41.
assessment criteria, making a total of 48 assessment parameters. Each individual parameter
is given a score from one (“Not working”) to four (“Works well”) and a weighting. The overall
score yielded by the accreditation matrix represents the weighted average of these
48 assessment parameters.
However, it is inappropriate to use the overall score yielded by the accreditation matrix 43.
to assess the PAs’ internal control systems in relation to legality and regularity of
39 As per the table on p. 9 of Annex 3 to Guideline No 2 on sampling methodology, for a high inherent risk and an estimated error of 10 %.
40 Article 7 of the CAP Horizontal Regulation.
41 Example adapted from DG AGRI’s Guideline No 3.
27
expenditure, as these depend mainly on only two of the 48 assessment parameters:
administrative checks and on-the-spot checks on the processing of claims, including their
validation and authorisation (under “Control activities”).
Under the framework proposed by DG AGRI, low scores on these two parameters may 44.
be compensated for by higher scores on others (e.g. information and communication,
monitoring). These other factors are not directly related to legality and regularity and should
not, therefore, be used to compensate in this way. For legality and regularity purposes, given
the overwhelming importance of these two parameters, such compensation may lead CBs to
overrate internal control systems and derive more assurance from them than is warranted.
Furthermore, this framework does not sufficiently take into account systemic weaknesses
previously reported both by DG AGRI and by us following audits of PAs and final
beneficiaries.
In three42 of the six Member States visited, we found that the CBs had not taken 45.
sufficient account of known weaknesses in their systems assessments. Box 1
Box 1 – Situations where the accreditation matrix did not produce reliable results for legality and
regularity purposes
below presents
the situation we found in one of these Member States (Germany (Bavaria)), along with the
opposite situation found in Romania, where the CB actually made full use of such known
weaknesses in its assessment. These situations support our view that the accreditation
matrix is an inappropriate tool for assessing the functioning of internal control systems for
legality and regularity.
In Germany (Bavaria), the accreditation matrix yielded a rating of “Works” for the state’s EAFRD
(non-IACS) internal control system. This rating was not consistent with the following evidence, which
suggested that the control system for such expenditure was weaker:
- for the 2013 and 2014 financial years, DG AGRI had calculated that, in relation to the legality
and regularity of expenditure, there was a material level of error, and
42 Germany (Bavaria), Italy (AGEA) and the United Kingdom (England).
28
- the CB’s own verifications showed that a significant number of transactions contained financial
errors (25 from the 2015 financial year, i.e. one third of the audited population).
In Romania, the accreditation matrix initially yielded a rating of “Works” for the internal control
systems for both funds (EAFRD and EAGF) and both strata (IACS and non-IACS). However, applying its
professional judgement using the information available, including known weaknesses previously
reported by DG AGRI and by us, the CB disregarded this outcome and reduced the corresponding
ratings to “Does not work” for EAGF (IACS) and “Works partially” for EAFRD (IACS) and EAGF (non-
IACS).
During its conformity clearance inquiries, DG AGRI uses key and ancillary controls 46.
(see Box 2
Box 2 – What are key and ancillary controls?
for definitions) to assess whether the internal control systems in place at Member
State level are capable of ensuring the legality and regularity of expenditure.
43
“Key controls shall be the administrative and on-the-spot checks necessary to determine the
eligibility of the aid and the relevant application of reductions and penalties”.
“Ancillary controls shall be all other administrative operations required to correctly process claims”.
Even though DG AGRI itself uses key and ancillary controls to assess whether PAs’ 47.
internal control systems ensure the legality and regularity of expenditure, it requires CBs to
use a different tool for the same purpose: the accreditation matrix, which is not fit for that
purpose.
The CBs’ sampling method for IACS transactions, based on the PAs’ lists of randomly
selected on-the-spot checks, entaileds a series of risks that were not overcome
As mentioned above the CAP Horizontal Regulation requires CBs to provide an opinion 48.
on the legality and regularity of the expenditure for which reimbursement has been
requested from the Commission during the financial year being audited. Therefore, when
43 Article 12 (6)(a) and (b) of Regulation (EU) No 907/2014.
29
designing an audit sample, the CBs ’shall consider the purpose of the audit procedure and
the characteristics of the population from which the sample will be drawn’44
DG AGRI guidelines
.
45
- Sample 1: drawn by CBs from the list of beneficiaries randomly
split the work for substantive testing into two samples: 49.
46
- Sample 2: drawn by CBs from all payments for the year in question. For Sample 2, CBs
need to re-perform only administrative checks, not on-the-spot checks (see
paragraphs 62 to 67).
selected by the PAs for
on-the-spot checks (see paragraph 6). For Sample 1 transactions, CBs are required to re-
perform both the on-the-spot checks carried out by the PAs and the full range of
administrative checks (receipt of the claim and eligibility checks, validation of
expenditure, including the authorisation, execution and recording in the accounts of the
corresponding payment).
The timetable of the CBs substantive testing and reporting to the Commission are 50.
described in Annex IV
Sample 1 transactions are chosen from a list of randomly selected on-the-spot checks 51.
performed by the PA on claims, which, if they are IACS, are very likely to lead to payments
during the corresponding financial year. Hence, the results of CBs’ testing not only serve the
purpose of expressing an opinion on the legality and regularity of expenditure, but also help
the CBs to form an opinion on the functioning of the PAs’ internal control systems, which is
one of the other audit objectives laid down in the CAP Horizontal Regulation.
.
The number of transactions upon which CBs re-perform all of the PA checks (see Table 3
44 Paragraph 6 of International Standard on Auditing (ISA) 530 – Audit sampling.
52.
below paragraph 64) is often considerably greater than the number of transactions usually
45 Section 6.2 of Guideline No 2 on sampling methodology.
46 The random sample represents between 20 to 25 % of PAs’ on-the-spot checks. The remaining PAs’ on-the-spot checks are sampled based on risk.
30
required for compliance testing on internal control systems47
For audit results to be reliable, it is important that all samples used for substantive 53.
testing be representative. In order to be representative, samples must be drawn from the
total population using statistically valid methods and remain unchanged. Because the CBs
extract their samples in part from the PAs’ samples, the results of their work can only be
representative if the PAs’ samples are themselves representative. Furthermore, any
replacement of transactions contained in the initial samples must be appropriately justified
and documented.
. This gives them greater
assurance that the PAs’ systems for administrative and on-the-spot checks comply with the
applicable regulations.
DG AGRI requires CBs to assess the representativeness of the PAs’ samples. However, in 54.
order to comply with such a requirement, both the PAs and the CBs need to maintain a
sufficient and reliable audit trail to confirm that the samples are representative, drawn from
the entire population and have remained unchanged.
DG AGRI requires Sample 1 IACS transactions for substantive testing to be selected 55.
based on the amounts claimed by farmers (claim-based selection) rather than the amounts
they are actually paid48
However, for this method to work as intended, close cooperation and communication 56.
between the PAs and the CBs is important. The CBs need to be updated regularly on checks
performed by the PA so that the CB can re-perform these checks shortly after the PA’s visit.
, so that the CBs can perform their re-verifications as soon as possible
after the PAs’ on-the-spot checks. The advantage of the claim-based selection is that the
conditions the CBs find when they inspect the selected farms on the spot will be very similar
to those found by the PAs.
47 Annex 3 of Guideline 2 recommends a minimum of 30 transactions for each Fund.
48 Section 6.7 of Guideline No 2 on sampling parameters.
31
We found that in three of the five Member States visited where the CB re-performed the
IACS on-the-spot checks49, this condition was not met (see Box 3
Box 3 – Delays in the CBs’ re-performance of IACS Sample 1 on-the-spot checks for the 2014 claim
year (2015 financial year)
).
In Poland, the CB completed the re-performance of 53 of the 60 EAGF IACS Sample 1 transactions
after the end of the 2014 calendar year, with some being re-performed as late as March 2015. This
was because the PA had been late in sending the CB the reports on its on-the-spot checks.
In Romania, the CB re-performed all six classical on-the-spot checks50
In Germany (Bavaria), up to three months had passed between the PA’s initial farm visits and the
CB’s on-the-spot re-verifications for EAGF IACS transactions, and up to ten months for EAFRD IACS
transactions.
for EAFRD IACS Sample 1
transactions on average six months after the PA’s checks.
Also, for this approach to be valid, the CBs need to have immediate access to the list of 57.
on-the-spot checks initially selected by the PAs, in order to ensure that this list is not
subsequently changed. Moreover, a CB must not be allowed to exclude transactions from its
sample as a result of delays in the PA’s decision granting the relevant aid. Allowing the CBs
to exclude such transactions from the sample increases the risk of PAs intentionally delaying
their decisions and payments for certain transactions, which might lead to errors in the CBs’
calculations. Such cases not only undermine the representativeness of the results and the
validity of their extrapolation to overall expenditure, but may also cause an incorrect
reduction in the error rate found by the CBs (see Box 4
49 Italy is not counted for this purpose because for the 2015 financial year the CB did not test IACS transactions on-the-spot because of its late appointment.
).
50 On-the-spot checks can take the form of classical on-the-spot visits (to agricultural holdings or investments subject to aid) or of remote sensing (review of recent satellite images of parcels).
32
Box 4 – Exclusion of transactions from IACS Sample 1 because the PA did not execute payment
until after the CB had completed its audit work
In Romania, the CB excluded five transactions from its EAFRD IACS Sample 1 after completing its
audit work because the PA had not yet executed the corresponding payments. Consequently, the CB
did not take into consideration the results of the tests on these transactions in calculating the overall
error rate upon which its opinion on the legality and regularity of transactions was based.
The application of the claim-based selection method also relies on the CBs not letting 58.
the PAs know which transactions they have selected until after the latter have already
performed their checks, so that the PAs do not know which transactions the CBs will later
scrutinise (the Commission has also recognised this risk in cases where CBs accompany the
PAs on their initial checks)51. However, we found that in Italy such safeguards were not in
place (see Box 5
Box 5 – The re-performance exercise was compromised in Italy because the CB had given the PA
advance notice of which beneficiaries would be subject to re-performance
).
In Italy, for the 2015 claim year (2016 financial year), the CB gave the PA the list of transactions it had
selected for Sample 1 IACS for both funds (EAGF and EAFRD) before the PA had carried out the
majority of its initial on-the-spot checks. As a result, the PA already knew before starting its on-the-
spot checks which transactions would subsequently be subject to re-performance by the CB.
If, when carrying out its on-the-spot checks, the PA knows which transactions are scheduled for re-
performance by the CB at a later date, it is likely to scrutinise these transactions more closely. As a
result, these checks will be more accurate than those not scheduled for re-performance, so the CB
will find fewer errors.
Moreover, since the results for these transactions will be used to extrapolate the level of error for
the entire population, these biased transactions will render the overall error rate unrepresentative
and more than likely result in it being understated.
51 Section 6.2 of Guideline No 2 on sampling methodology.
33
A portion of the non-IACS transactions upon which the CBs perform substantive testing is
not representative of expenditure in the financial year audited
For non-IACS transactions (both EAGF and EAFRD), there is a significant disparity 59.
between:
- the period for which the on-the-spot checks are reported, which is the calendar year
(from 1 January to 31 December 2014 for the 2015 financial year, in this case), and
- the period for which expenditure is paid, which for the 2015 financial year is from
16 October 2014 to 15 October 2015.
As a result, some of the beneficiaries subject to on-the-spot checks performed during 60.
the 2014 calendar year were not reimbursed in the 2015 financial year52
As a result, the number of transactions the CBs use in calculating the overall error rate is 61.
lower, thus reducing its accuracy and, in turn, affecting the reliability of their opinions on the
legality and regularity of expenditure for the financial year.
. The CBs cannot use
the results of such transactions in their calculation of the error rate for the financial year
concerned. Such results may only be used to express an opinion on the functioning of the
internal control system and provide a statement on the assertions in the MD and its
underlying control statistics.
Substantive testing
For most transactions, the guidelines do not require the CBs to perform on-the-spot
testing at final beneficiary level
According to internationally accepted audit standards, ‘in performing a reasonable 62.
assurance audit, public sector auditors gather sufficient appropriate audit evidence to
52 The beneficiaries reimbursed before 16 October 2014 were recorded in the previous financial year and those (exceptional cases) reimbursed after 15 October 2015 were recorded in the following financial year.
34
provide a basis for the auditors’ conclusions’53. They should use a variety of techniques for
this purpose, such as observation, inspection, inquiry, re-performance, confirmation and
analytical procedures. Furthermore, under EU law, the CBs’ substantive testing of
expenditure must cover verifying the legality and regularity of the underlying transactions at
final beneficiary level54
Inspection involves examining books, records and other elements (such as the 63.
beneficiaries’ right of use for the land in question, fertilisation registers, building permits,
etc.) or physical assets (such as land, animals, equipment, etc.). Inquiry involves obtaining
information and explanations from final beneficiaries, in the form of either written
statements or more informal oral discussions.
.
As mentioned above, for sample 2, the Commission’s guidelines only require CBs to re-64.
perform administrative checks (see paragraph 49). In Table 3
53 Section 7.1 of ISSAI 4200 - Compliance Audit Guidelines. Compliance Audit Related to the Audit of Financial Statements.
, we calculated the proportion
of Sample 2 transactions in the total sample for the six Member States visited.
54 Article 7(3) of Regulation (EU) No 908/2014.
35
Table 3 – Proportion of Sample 2 transactions used for substantive testing by the six Member States visited
Member State EAGF EAFRD Sample 1 Sample 2 Total
sample % of Sample 2
in total Sample 1 Sample 2 Total
sample % of Sample 2
in total (1) (2) (3) = (1)+(2) (4) = (2)/(3)*100 (5) (6) (7) = (5)+(6) (8) = (6)/(7)*100
Germany (Bavaria) 30 41 71 58 % 60 146 206 71 % Spain (Castilla y Léon) 52 123 175 70 % 50 122 172 71 % Italy (AGEA1) 30 192 222 86 % 30 91 121 75 % Poland 90 58 148 39 % 97 33 130 25 % Romania 161 292 453 64 % 140 245 385 64 % United Kingdom (England) 30 40 70 57 % 59 26 85 31 % Average proportion of Sample 2 in the Member States/Regions visited (with no on-the-spot testing)
62 % 56 %
1 The ‘Agenzia per le erogazioni in agricoltura’ (AGEA) is one of the 11 Italian PAs.
36
As the table above shows, in four of the Member States/Regions visited, Sample 2, for 65.
which no on-the-spot testing is done, accounted for the majority of the overall sample used
by CBs for substantive testing.
For sample 2, gathering evidence based only on a re-performance of administrative 66.
checks at PA level, will often not provide the CBs with sufficient and appropriate audit
evidence as required by ISSAI 4200. By doing so, the CBs are not using two audit collection
methods that are very important in the context of the CAP expenditure: inspection and
inquiry. Without these methods, important audit evidence cannot be collected, such as
evidence that farmers are using their land as stated in their declarations and proof of the
existence of physical assets purchased through investment projects. This approach therefore
falls short of internationally accepted audit standards.
Without sufficient on-the-spot testing, the CBs may not be able to obtain reasonable 67.
assurance that the relevant expenditure is legal and regular: as we have demonstrated in our
annual reports, most of the errors in CAP expenditure (particularly for IACS) are found on the
spot55
The guidelines require the CBs only to re-perform PAs’ initial checks, rather than carry out
all the audit procedures the CBs consider necessary to obtain reasonable assurance
.
‘The nature, timing and extent of procedures performed […] are determined by public 68.
sector auditors applying professional judgement’56
DG AGRI required CBs, for their substantive testing, only to re-perform (re-verify) the 69.
PAs’ initial checks, for both Sample 1 and Sample 2 transactions
. With regards to the standards applicable
when performing the reasonable assurance audit see also paragraph 62.
57
55 See, for example, paragraphs 7.19 and 7.26 of the Court’s annual report for the 2014 financial year (OJ C 373, 10.11.2015) and paragraphs 7.14, 7.18 and 7.22 of the Court’s annual report for the 2015 financial year (OJ C 375, 13.10.2016).
. For example, if the initial
56 Paragraph 21 of ISSAI 4200.
57 Section 6.2 of Guideline No 2 on sampling methodology.
37
on-the-spot checks had been done by the PAs using remote sensing58, then the CBs were
always required to re-perform them using the same method59
Re-performance is defined by internationally accepted audit standards
.
60
While re-performance is a valid audit evidence collection method, the Commission 71.
should not, in our opinion, require the CBs to limit themselves to such method for all
transactions subject to substantive testing but instead leave it up to the CBs to determine
the extent of its use. CBs might choose and perform audit steps and procedures that they
themselves consider appropriate, which should not be limited to re-performance.
Substantive testing based only on re-performance might not allow the CBs to obtain
sufficient audit evidence to form a reasonable assurance opinion.
as 70.
‘independently carrying out the same procedures already performed by the audited entity’,
in this case the PA. For example, CBs may re-perform administrative checks to test whether
the PAs made the correct decisions when granting aid.
Conclusion and audit opinion
The guidelines require the CBs to calculate two different error rates and the use made of
these rates by the CBs and DG AGRI is not appropriate
In order to comply with DG AGRI guidelines61, CBs must calculate two error rates for 72.
legality and regularity purposes (see Table 4
58 Computer-based checks and detailed photo interpretation of ortho-images performed on recent high-resolution satellite or aerial images of parcels.
):
59 Section 9.2 of Guideline No 2 on the performance of on-the-spot checks.
60 Section 7.1.5 of ISSAI 4200.
61 Glossary of Guideline No 2.
38
Table 4 - The error rates to be calculated in accordance with DG AGRI guidelines
Characteristics Error rate (ERR) ‘Incompliance’ rate (IRR) Use made by the CB To express an opinion on the
completeness, accuracy and veracity of the PA’s annual accounts, on the proper functioning of its internal control system and on the legality and regularity of the expenditure for which reimbursement has been requested from the Commission.
To assess whether their examination puts in doubt the assertions made in the MDs, including the errors reported by PAs in the control statistics.
Population to which the rate refers to
Expenditure for the financial year (e.g. 16 October 2014 to 15 October 2015).
Controls performed by the PA during the calendar year (e.g. 1 January 2014 to 31 December 2014).
PA procedures taken into account
The entire control process performed by the PA, from receipt of claims until payment and accounting.
Only PA’s primary eligibility controls (administrative and on-the-spot checks), before sanctions are applied.
Errors taken into account Only overpayments by the PA are taken into account.
It takes into account both the over- and under-validations by the PA.
Article 9 of the CAP Horizontal Regulation requires the CBs to express a “limited” 73.
(negative) assurance conclusion62
Thus, in order to be compliant with internationally accepted audit standards, the 74.
conclusion on the MD can be drawn from the results of the audit work done in relation to
the reliability of accounts, the proper functioning of the internal control system and the
legality and regularity of transactions (including the ERR), i.e. as a “by-product” of this work.
in relation to the MD. Such a conclusion does not require
fully-fledged audit work, but merely a review, usually limited to analytical procedures and
enquiries. Tests of details, such as substantive testing of transactions, are not required. A
limited assurance review therefore provides a lower level of assurance than a reasonable
assurance audit. The limited assurance conclusion is usually worded as follows: “nothing has
come to our attention that would indicate that the subject matter is not in compliance, in all
material respects, with the stated criteria”.
62 For a more detailed definition of limited assurance and how it differs from reasonable assurance, see paragraphs 20 and 21 of ISSAI 4200.
39
However, DG AGRI guidelines additionally require the CBs to calculate the IRR in order to
draw a conclusion on the MDs.
Furthermore, in its 2015 annual activity report63
The IRR does not represent the level of error in the expenditure, but rather the financial 76.
impact of weaknesses in the PA’s administrative and on-the-spot checks.
, DG AGRI explained that, in principle, it 75.
intended to use the IRR to estimate an adjusted error rate for legality and regularity
purposes. So, whereas the CBs’ opinions on the legality and regularity of expenditure are
based on the ERR, DG AGRI‘s assurance model uses a different indicator of error, the IRR, in
order to estimate the amounts at risk from a legality and regularity perspective.
DG AGRI’s guidelines require the CBs to use the ERR calculated for legality and regularity 77.
purposes to also form an opinion on the completeness, accuracy and veracity of the PAs’
annual accounts (see Table 4
However, this is contrary to the CAP Horizontal Regulation
). In its annual review of the PAs’ accounts (the “financial
clearance procedure”), DG AGRI itself also uses the error rate for legality and regularity in
order to assess the reliability of the PAs’ accounts.
64
63 Explanatory Box: Annex 10-2.1.
which, for the financial 78.
clearance procedure, clearly separates the issue of the reliability of the PAs’ accounts from
that of the legality and regularity of expenditure. This distinction is appropriate because, for
example, a payment may be properly accounted for despite not being legal and regular (e.g.
because a farmer claimed aid for land worked by another farmer); and, vice-versa, a
payment may be legal and regular despite being incorrectly accounted for (e.g. recorded as
reimbursement of an investment project instead of as annual area-based payment). Given
this separation, DG AGRI’s use of the ERR to assess the reliability of the PAs’ accounts is
inappropriate.
64 Article 51 and recital 35 of the CAP Horizontal Regulation, which states that “The decision of the clearance of accounts should cover the completeness, accuracy and veracity of the accounts but not the conformity of the expenditure with Union law”.
40
The CBs’ opinion on the legality and regularity of expenditure is based on an understated
total error
The two indicators of error (ERR and IRR) are calculated as the difference between what 79.
the CB considers eligible and what the PA has previously validated. The differences between
what beneficiaries claimed and what the PA validated after its on-the-spot checks, which
constitute the errors reported by the PAs in the control statistics, are not taken into
consideration in the CBs’ ERR.
As paragraph 6 states, as a general rule, only 5 % of claims are subject to on-the-spot 80.
checks by the PAs. And yet, for the six Member States visited, the average share of
transactions subject to PA on-the-spot checks in the overall sample audited by CBs
(see Table 3) is 38 % for EAGF and 44 % for EAFRD. Hence, such transactions are
overrepresented in the overall CB samples. Figure 3
Figure 3 – The proportion of transactions subject to PA on-the-spot checks in the EAGF
population and in the average CB sample
below illustrates this situation with the
average figures for the EAGF fund.
Source: European Court of Auditors.
Where transactions have previously been subject to on-the-spot checks, the PA has 81.
already detected and reported errors for these in the control statistics. These errors will no
longer be identified by the CBs when comparing their results to the PAs’ results. However,
EAGF population
Transactions subject to PA on-the-spot checks
Transactions subject only to PA administrative checks
Sample 1
Sample 2
41
potential errors will remain uncorrected for the 95 % of the population not subject to on-
the-spot checks.
Thus, the CBs must accurately reflect the proportion of Sample 1 transactions that have 82.
previously been subject to on-the-spot checks ensuring a true representation of the
characteristics of the total population. Therefore, they need to add to their own error rate
the PA’s error rate, from the control statistics, for its random on-the-spot checks on the
share of Sample 1 transactions in excess of 5 %65
The PA’s error rate from the control statistics likewise needs to be added to the CB’s 83.
error rate for those Sample 2 transactions which have not been subject to any on-the-spot
checks, either by the PA
that have previously undergone such
checks.
66
This adjustment is needed because CBs are required to express an opinion on the 84.
legality and regularity of the entire population of payments, and not only on the
effectiveness of the PAs’ procedures for on-the-spot checks. In the absence of such
adjustments, the ERR is likely to be significantly understated for both Sample 1 and
Sample 2.
or by the CB.
In addition to the lack of such adjustments, the weaknesses reported in paragraphs 56, 85.
57, 58, 67 and 71 of this report are also likely to result in the error rate used by CBs for their
opinion on legality and regularity being understated.
CONCLUSIONS AND RECOMMENDATIONS
With a budget of 363 billion euro (in 2011 prices) for the 2014-2020 programme period, 86.
CAP expenditure is under shared management between the Commission and the Member
States. While the Commission delegates the implementation of the budget to PAs
65 5 % is the general rule, but where PAs have verified on-the-spot a different percentage of beneficiaries, that rate should be taken into consideration instead.
66 Given, however, that Sample 2 is randomly drawn from the list of payments, 5 % of this sample can be expected to have undergone an on-the-spot check by the PA.
42
designated by Member States, it retains ultimate responsibility and is required to ensure
that the EU’s financial interests are protected to the same standard as though the
Commission were performing the delegated budget implementation tasks itself.
The Commission draws assurance that expenditure is legal and regular from three layers 87.
of information: the checks performed by the PAs, the audit work carried out by the CBs and
the results of the Commission’s checks thereon. When such assurance is considered
insufficient, the Commission exercises its ultimate responsibility for legality and regularity
through its own conformity clearance procedures which may result in financial corrections.
The CBs have been entrusted with their role as the PAs’ independent auditors since 88.
1996. While they were initially responsible for issuing a certificate on the reliability of
accounts and on internal control procedures, their responsibilities were increased for the
2014-2020 period. CBs have been required, since the 2015 financial year, to also provide an
opinion, drawn in accordance with internationally accepted audit standards, stating whether
the expenditure for which reimbursement has been requested from the Commission is legal
and regular. The 2015 financial year, then, was the first for which the Commission could use
the CBs’ enhanced work on legality and regularity in preparing its AAR.
In this context, the present report assessed whether the CBs’ new role was a step 89.
towards a single audit model and if the Commission took due account of it in its assurance
model. We also assessed whether the framework set up by the Commission enabled the CBs
to draw an opinion on the legality and regularity of CAP expenditure in accordance with the
applicable EU regulations and internationally accepted audit standards.
We conclude that although the CBs’ new role is a positive step towards a single audit 90.
model, the Commission could take very limited assurance from the CBs’ work on legality and
regularity . In addition, we conclude that the framework designed by the Commission for the
first year of implementation of the new work of the CBs has significant weaknesses. As a
result, the CBs’ opinions do not fully comply with the applicable standards and rules in
important areas.
We consider the CBs' new role in relation to the legality and regularity of expenditure to 91.
be a positive step towards a single audit model, in which the different layers of control and
43
audit can be complementary, thus avoiding uncoordinated, overlapping controls and audits.
This is because the CBs' outputs have the potential to help Member States to strengthen
their control systems, reduce audit and control costs and enable the Commission to obtain
independent additional assurance as to the legality and regularity of expenditure.
Furthermore, the CB's work on legality and regularity can also help the Commission in
improving how it estimates the overall residual error, assessing the need for conformity
audits, determining more accurately and completely the amounts to be excluded from the
EU budget, making more extensive use of extrapolated financial corrections and reviewing
the accreditation status of PAs (paragraphs 22 to 28).
Against this backdrop, we analysed the Commission’s existing assurance model and the 92.
changes introduced to take account of the CBs’ enhanced role. We noted that the
Commission’s assurance model remains based on the Member States’ controls results. For
the 2015 financial year, the CBs' opinions on legality and regularity were merely one factor
to be taken into account when the Commission calculated its adjustments of the errors
reported in the Member States' control statistics. For the financial years 2014 and 2015, the
Commission made top-ups which increased the overall residual error by between two to
four times what was reported by Member States. The CBs’ opinions on legality and regularity
are the only source which provides independent assurance on legality and regularity on an
annual basis. Thus, once the CBs’ work is done in a reliable manner this independent
assurance should in our view become the key element for the Commission’s Director-
General for DG AGRI when assessing whether expenditure is legal and regular
(paragraphs 29 to 36).
Recommendation 1 - Assurance from the CBs’ work on legality and regularity
The Commission should use the CBs’ results, when the work is defined and performed in accordance
with the applicable regulations and internationally accepted audit standards, as the key element of
its assurance model regarding the legality and regularity of expenditure.
We then examined the extent to which the Commission's guidelines complied with the 93.
applicable regulations and internationally accepted audit standards at the different stages of
the audit process. For the risk assessment procedure, we observed that the Commission
required CBs to use the accreditation matrix in order to determine the extent to which they
44
should rely on the PAs' internal control systems. The accreditation matrix is used by CBs to
assess the PAs' compliance with accreditation criteria and contains 48 assessment
parameters, out of which only two have a significant impact on the legality and regularity of
expenditure: the administrative and the on-the-spot checks on the processing of claims.
Thus, it is inappropriate to use this matrix for legality and regularity purposes, because 94.
this may increase the level of assurance the CBs derive from PAs’ internal control systems
and decrease the size of their samples for substantive testing. During its conformity
clearance inquiries, the Commission uses a different tool to assess the effectiveness of the
PAs internal control systems: a list of key and ancillary controls, which comprise the
administrative checks and physical on-the-spot checks, as well as other administrative
operations, required to ensure the correct calculation of the amounts to be paid
(paragraphs 38 to 47).
Recommendation 2 – Risk assessment focused on key and ancillary controls
The Commission should revise its guidelines so that the CBs’ risk assessment as regards legality and
regularity is focused on the key and ancillary controls already used by the Commission,
complemented by any other evidence that the CBs judge appropriate in accordance with
internationally accepted audit standards.
For the sampling of transactions, the Commission required CBs to use two samples 95.
whose results are then merged: Sample 1, drawn from the list of beneficiaries randomly
selected by the PAs for on-the-spot checks, and Sample 2, drawn from all payments for the
financial year in question. This approach allows the CBs to test a significant number of
transactions (Sample 1) which have also been verified on-the-spot by the PAs, so that the
CBs gather evidence regarding the functioning of the PAs internal control system and the
reliability of the PAs control statistics where the results of the PAs on-the-spot checks are
reported. However, because the CBs extract Sample 1 from the PAs’ random on-the-spot
checks, their work can only be representative if the samples initially selected by the PAs are
themselves representative.
Furthermore, for IACS transactions, which are mainly annual area-based payments, 96.
Sample 1 was selected based on the amounts claimed by farmers, before the PAs had
45
determined the amounts to be paid. The Commission chose this solution so that the CBs
could perform their re-verifications as soon as possible after the PAs’ checks, when the
conditions the CBs found when inspecting the selected farms on-the-spot would be very
similar to those found by the PAs. However, this solution entails several risks, such as:
insufficient cooperation between the PAs and CBs which significantly delays the timing of
the CBs testing, the risk of transactions potentially affected by error being replaced in the CB
sample or the risk of the CBs announcing their visit before the PAs carry out their initial
checks (paragraphs 48 to 58).
Recommendation 3 – Safeguards for IACS sampling based on PAs’ on-the-spot checks
For the selection by the CBs of IACS transactions from the list of claims randomly selected by the PAs
for on-the-spot checks, the Commission should reinforce its guidelines, requiring CBs to put in place
appropriate safeguards that:
- ensure that the CBs’ samples are representative, are transmitted upon request to the
Commission and that an appropriate audit trail is maintained. This means that the CBs need to
check whether the Pas’ samples are representative;
- enable the CBs to plan and carry out their visits shortly after the PAs have carried out their on-
the-spot checks;
- ensure that CBs do not disclose their sample to the PAs before the latter carried out their on-
the-spot checks.
For non-IACS expenditure, which is mainly reimbursement-based expenditure for 97.
investments in farms and rural infrastructure, there is a disparity between the period for
which PAs’ on-the-spot checks are reported (e.g. from 1 January to 31 December 2014 for
the 2015 financial year) and the period for which expenditure is made (e.g. from
16 October 2014 to 15 October 2015 for the 2015 financial year). This means that those
payments executed before 16 October 2014 or after 15 October 2015 will not represent
expenditure for the 2015 financial year (paragraphs 59 to 61).
46
Recommendation 4 – Non-IACS sampling based on payments
The Commission should revise its guidelines with regards to the sampling method for non-IACS
expenditure, so that the CBs sample non-IACS expenditure directly from the list of payments
executed during the financial year being audited.
For Sample 2, the Commission required CBs to carry out only a re-performance of the 98.
PAs' administrative checks. In most Member States visited, Sample 2 accounted for the
majority of the total CB sample used for substantive testing. However, gathering evidence
based only on a documentary review at PA level, without any on-the-spot testing at final
beneficiary level, will often not provide the CBs with sufficient and appropriate audit
evidence as required by internationally accepted audit standards, as it deprives the auditor
of two audit evidence collection methods that are very important in the context of the CAP
expenditure: inspection and inquiry (paragraphs 62 to 67).
The Commission required CBs, for their substantive testing, only to re-perform (re-99.
verify) the PAs’ initial checks, for both Sample 1 and Sample 2 transactions. Re-performance
is defined by internationally accepted audit standards as ‘independently carrying out the
same procedures already performed by the audited entity’, in this case the PA. However,
these standards state that auditors should choose and perform all audit steps and
procedures that they themselves consider appropriate. As such, because the Commission
limits the CBs to re-performance, their substantive testing is incomplete and does not
provide them with sufficient audit evidence to form a reasonable assurance opinion
(paragraphs 68 to 71).
Recommendation 5 – Substantive testing carried out on-the-spot
The Commission should revise its guidelines to allow the CBs to carry out:
- on-the-spot testing for any transaction audited;
- all audit steps and procedures that they themselves consider appropriate, without being limited
to re-performing the PAs’ initial checks.
47
The Commission requires the CBs to calculate two indicators of error in relation to 100.
legality and regularity:
- the ERR, which is based on the CB’s audit of the PA’s entire control process, from
receipt of claims until payment and accounting, and is used to express an opinion on the
legality and regularity of expenditure; and
- the IRR, which measures only the financial impact of errors in the PA’s primary eligibility
controls (administrative and on-the-spot checks) before sanctions are applied and is
used to assess whether the CBs’ examination casts doubt on the assertions made in the
MDs, including the levels of error reported by PAs in the control statistics.
However, the CAP Horizontal Regulation requires only limited assurance regarding the MDs,
which is usually based on work that is limited to analytical procedures and enquiries, with no
tests of details.
The IRR is not necessary for ensuring compliance with internationally accepted audit 101.
standards, as the conclusion on the MDs can be drawn as from the results of the audit work
done in relation to the reliability of accounts, the proper functioning of the internal control
system and the legality and regularity of transactions. Furthermore, the Commission
envisaged using the IRR instead of the ERR in the calculation of the adjusted error rate,
despite the fact that the IRR does not represent the level of error in the expenditure but
rather the financial impact of weaknesses in the PA’s administrative and on-the-spot checks.
On the other hand, however, the ERR has been wrongly used by both the CBs and the
Commission to form a view on the completeness, accuracy and veracity of the PAs' accounts
(paragraphs 72 to 78).
48
Recommendation 6 – A single error rate in relation to legality and regularity
The Commission should revise its guidelines so that they do not require the calculation of two
different error rates regarding legality and regularity. A single error rate, based on which the CBs
expressed their reasonable assurance opinion on legality and regularity of expenditure and limited
assurance conclusion on the assertions made in the management declaration, would satisfy the
requirements of the CAP Horizontal Regulation.
The Commission and the CBs should not use such an error rate for legality and regularity to judge
the completeness, accuracy and veracity of the PA’s annual accounts.
The ERR is calculated as the difference between what the CB considers eligible and what 102.
the PA has previously validated. The differences between what beneficiaries claimed and
what the PA validated after its on-the-spot checks, which constitute the errors reported by
the PAs in the control statistics, are not taken into consideration in the ERR which is the basis
of the CBs’ opinion on legality and regularity of expenditure. Where transactions have
previously been subject to on-the-spot checks, the PAs have already detected and reported
the corresponding errors in their control statistics. These errors will no longer be identified
by the CBs when comparing their results to the PAs’ results, and the CBs are therefore likely
to find fewer errors for such transactions. Thus, the more transactions have previously been
subject to PA checks in the CB sample, the lower the overall level of error found by the CBs
will be.
In the six Member States visited, the average share of transactions subject to PA on-the-103.
spot checks in the overall sample audited by CBs was 38 % for EAGF and 44 % for EAFRD.
However, as a general rule, the percentage of transactions in the total population subject to
PA on-the-spot checks is only 5 %. Thus, in the CB samples, the transactions which had
previously been subject to PA on-the-spot checks – and which were therefore less affected
by error in CB testing –were overrepresented in the overall CB samples. This led to
underestimated levels of error being reported by the CBs, because the CBs had not made
any adjustments to ensure a faithful representation of the much smaller proportion of
transactions previously subject to PA on-the-spot checks in the actual population
(paragraphs 79 to 85).
49
Recommendation 7 – An error rate representative of the population covered by the audit opinion
The Commission should revise its guidelines so that:
- for IACS transactions which are sampled from the list of PAs’ random on-the-spot
checks, the overall error calculated by the CBs also includes the level of error reported
by the PAs in the control statistics, extrapolated to the remaining transactions not
subject to PA on-the-spot checks .The CBs have to ensure that the control statistics
compiled by the PAs are complete and accurate;
- for the transactions sampled by the CBs directly from the full population of payments,
no such adjustment is necessary because the sample would be representative of the
underlying populations audited.
The above recommendations 2 to 7 have as their implementation target the next 104.
revision of the Commission's guidelines applicable from the financial year 2018 onwards. The
Commission finalised in January 2017 new guidelines to be applied by CBs from 2018
financial year. These revised guidelines have not been subject to our review (see
paragraph 17).
This Report was adopted by Chamber I, headed by Mr Phil WYNN OWEN, Member of the
Court of Auditors, in Luxembourg at its meeting of 22 March 2017.
For the Court of Auditors
Klaus-Heiner LEHNE President
1
ANNEX I
The Commission’s shared management model, as presented in its 2015 AAR
Accredited Paying AgencyAdministrative checks on all aid applications by 15/05 n-1)
Selection of random and risk based samples for on-the-spot checks (OTSC)Performance of OTSC
Authorisation and payment of the aid (12/n-1 to 6/n)
Beneficiaries
Lodging of aid claims to
paying agency
Payment of aid to the
beneficiary (12/n-1 to
6/n)
On-the-spot-checks by PA on sample of
aid claims
Certification BodyOpinion
- on the annual accounts of the PA- on the proper function of the internal control
system in the PA- on the legality and regularity of the expenditure- whether the examination carried out by CB puts
in doubt the assertions made by the PA's director in his/her Management Declaration
CB opinion on PA report
CB reperforms and reverifies administrative and OTSC on a
sample of individual
transactions
European Commission : DG AGRIVerification of and reimbursement of monthly and quarterly payment declarations
Verification and clearance of annual accountsDG AGRI systems audits and conformity clearance procedure
Net financial corrections
Monthly (EAGF) and Quarterly (EAFRD)
declarations; Annual
Accounts,Control
Statistics,Management Declaration
DG AGRI conformity clearance systems audits to
Paying Agency
CB report & opinion
Reimburse-ment of
expenditure declarations;
Annual financial
clearance;Conformity clearance decisions;
Net financial corrections
1
ANNEX II
The Commission’s assurance model for legality and regularity of expenditure, as presented in its 2015 AAR
Accredited Paying Agency (PA)
Ex-ante administrative checks on 100 % of aid claimsEx-ante risk and random on-the-spot-checks (OTSC)
Reduction of aid claimsPayment of reduced amounts
aid applications
payment of aid
MS reporting
output
CB reperforms and reverifiesadministrative and OTSC on a
sample of individual transactions
DG AGRI conformity audits and net financial corrections
Reports to DG AGRI
Opinion sent to DG AGRI
Chec
kson
PA
DG AGRI certification audits
DG AGRI's accreditation audits
DG AGRI financial clearance
administrative and OTSC
Shared Management – CAP assurance model for legality and regularity
Where does DG AGRI get its assurance?
Architecture of MS management structure- accredited PAs
- independent audit bodies (CBs)
DG AGRI's conformity audits
DG AGRI's accreditation audits
DG AGRI audit and management opinions
Management Declaration of PAs
Opinions of the CBs
European Court of Auditors' findings
Net financial corrections
Ove
r 8 m
illio
n be
nefic
iarie
s
Annual AccountsAnnual Management Declaration
Results of controls carried out
Certification Body (CB)
Opinion- on the annual accounts of the PA
- on the proper functioning of the internal control system in the PA- on the legality and regularity of the expenditure
- whether the examination carried out by CB puts in doubt the assertions made by the PA's director in his/her Management Declaration
CB opinion onPA reports
1
ANNEX III
An example of an accreditation matrix, adapted from DG AGRI’s Guideline No 3 for the certification audit of the accounts
S T at 10 % or 15 %
S T at 5 % S T at 5 % S T at 50 % S T at 5 % S T at 10 % S T at 10 % S T at 5% W T
Administrative controls
4 0.40 4 0.20 2 0.10 3 1.50 3 0.15 3 0.30 3 0.30 4 0.20 20% 0.00 0.63
On-the-spot controls
4 0.40 4 0.20 2 0.10 2 1.00 3 0.15 3 0.30 2 0.20 4 0.20 20% 0.00 0.51
4 0.60 4 0.20 3 1.50 3 0.15 3 0.30 3 0.30 4 0.20 15% 0.00 0.49
4 0.60 4 0.20 3 1.50 3 0.15 3 0.30 3 0.30 4 0.20 15% 0.00 0.49
4 0.60 4 0.20 3 1.50 3 0.15 3 0.30 3 0.30 4 0.20 10% 0.00 0.33
4 0.60 4 0.20 2 1.00 3 0.15 3 0.30 3 0.30 4 0.20 20% 0.00 0.55
General conclusion 2.99
Assessment of the internal control system Works
Debts management
Advances and securities
Accounting
Execution of payments
Weighting/ Scoring
Claims processing, including Validation and autorisation
Internal audit
MATRIX - Assessment of the ICS for the IACS population
Assessment
Component procedure
Internal environment Control activities
Information and communication
Monitoring Evaluation at assessment
criteria
General conclusion
Organisation Human resources
Delegation Communication Information systems and security
Ongoing monitoring
1
ANNEX IV
The timetable of the PAs’ and CBs’ substantive testing and reporting to the Commission
5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6
PA inspections of IACS (EAGF and EAFRD) applicants, administrative
cross-checks run from claim submission
Control statistics (claim year 2014)
sent to the Commission Accounts and
management declaration
FY 2015 submitted to the
Commission (15 Feb)
PA work
Majority of payments to IACS scheme applicants
(advances paid from 16 Oct 2014; final payments from 1 Dec 2014)
PA samples of applicants to be
inspected are being drawn
Non-IACS population (EAGF and EAFRD)Receipt of applications, administrative and
on-the-spot checks, payments
Deadline for area-related aid applications (EAGF
and EAFRD)(15 May)
2014 2015 2016
CB re-performances of PA IACS inspections on 2014 applications
(EAGF and EAFRD)
CB should draw a random
sample of IACS inspections to
be re-performed
CB re-performances of PA non-IACS inspections(EAGF and EAFRD)
CB checks on payments made (re-performance of administrative checks) for IACS and non-IACS population
(EAGF and EAFRD)
CB reports for FY 2015 to be
submitted to the Commission
(15 Feb) CB work on PA
COM audit visits to MSs (CBs)
DG AGRI AAR for 2015 issued
(30 Apr)COM
work on CBs
Claim year 2014
Financial year 2015
2
REPLIES OF THE COMMISSION TO THE SPECIAL REPORT OF THE EUROPEAN
COURT OF AUDITORS
"THE CERTIFICATION BODIES' NEW ROLE ON CAP EXPENDITURE: A POSITIVE
STEP TOWARDS A SINGLE AUDIT MODEL BUT WITH SIGNIFICANT WEAKNESSES
TO BE ADDRESSED"
EXECUTIVE SUMMARY
Common Commission's reply to paragraphs II-IV:
The management and control system of the CAP expenditure is following a single control and audit
approach that is integrated in a pyramid of controls where each upper layer builds its work upon the
results of the previous layer, and where each layer may use the results of the upper layer to improve
its own controls. It is a dynamic model aiming not only at estimating an error rate but also aiming at
detecting the sources of errors, implementing remedial actions and reducing the error rates year
after year. For instance, the Integrated Administration and Control System (IACS) which governs
around 86 % of CAP expenditure is a good system for preventing and reducing the level of errors.
In the CAP assurance model applied by the Commission the independent CBs are a layer of the
pyramid of control. The CBs deliver every year an opinion on the PA's accounts and systems and,
from financial year 2015, on legality and regularity of expenditure. The Commission provides them
with guidelines.
DG AGRI auditors check the reliability of the opinions of the CBs on legality and regularity, in
order to assess whether they can be used as a source of assurance for the EU expenditure covered.
DG AGRI auditors also perform system audits which result where necessary, i.e. assurance not
obtained, in remedial action plans and net financial corrections, as well as interruptions, reductions
or suspensions of payments when the MS concerned does not take remedial actions.
The CAP single audit model
The CAP assurance model results in lower error rates and is cost effective. The overall error rate for
Natural resources (where the CAP represents 98 % of the expenditure) established by the Court for
financial year 2015 was 2.9 %.
V. The CAP assurance model already follows a "single audit" approach as it is an effective control
framework in which each layer builds on the assurance provided by others. This single audit model
gives reliable assurance on the legality and regularity of expenditure when taking into account all
corrective actions.
The new work of CBs reinforces assurance on the management and control system and on the
legality and regularity of the expenditure as:
3
- it creates a shorter circuit between the PA and its new first auditor, favouring quicker remedial
actions where necessary;
- it allows the Commission to better identify the expenditure at risk and focus its own audit and
corrective actions on it.
Where the CBs’ work cannot be relied on, DG AGRI continues to obtain independent assurance on
legality and regularity of expenditure from its own audits.
VI. To ensure continuity, the guidelines for 2015 – first year for the enhanced work of the CBs –
were elaborated on the basis of guidelines used for previous years' financial clearance. This was
also reflected in the integrated approach referred to in the new legislation. To take account of the
experience from the first 2015 exercise, the guidelines to CBs have already been revised. New
guidelines were finalised and presented to the MSs on 19 January 2017, for application from FY
2018, i.e. the earliest possible year.
(a) A new matrix has been introduced in the new guidelines for FY2018 for better reflection of the
level of assurance from the PA's internal control systems.
(b) The random samples controlled by PAs shall be selected in accordance with detailed EU rules
and guidance that include representativeness. The role of the CBs is to check that these rules have
been respected by the PAs and to extract second level samples that must be representative for the
random sample controlled by PAs. The issue for a portion of the sampling for the non-IACS
expenditure is addressed in the new guidelines for FY2018.
(c) In its single control and audit approach and considering the effectiveness of IACS, the
Commission considers that the audit work of the CBs should be focused on checking, by re-
performing them, the reliability of the controls done by the previous level in the pyramid of
controls.
(d) Re-performance is an audit methodology which, in accordance with IAAS, can be used to form
an audit opinion on legality and regularity. The Commission considers, as it is set out in the
guidelines, that, in the CAP framework, re-performance is sufficient to form an opinion on legality
and regularity of the expenditure. Furthermore, by definition, guidelines should never be understood
as mandatory instructions preventing those to whom they are addressed from taking their own
responsibilities. The guidelines should be seen as the minimum that is necessary to satisfy the
Commission, and not as a limit for CBs to do more.
Common Commission's reply to (e) and (f):
The Commission considers that the issues of the calculation of different error rates for legality and
regularity are addressed in the new guidelines for FY2018, taking into account Article 9(2)(b) of
Regulation (EC) No 1306/2013 which provides that, where appropriate, CBs may use an integrated
sample for both legality and regularity of expenditure and reliability of the accounts.
An incompliance rate (IRR) (cf. paragraph 74 for further details), to be added to the error rate found
by PAs and certified by CBs, is sufficient to form an audit opinion based on controls done by
previous layer of controls.
VII.
(a) The Commission considers that, whereas the guidelines for FY2015 can be improved, where
they are correctly applied they already in their current form enable the CBs to give an opinion on
legality and regularity in compliance with applicable rules and standards.
Where the audit work of the CBs is done in accordance with the applicable regulations and
guidelines their opinion are an important and valuable additional building block and should be the
key element for the Commission's assurance. This is in line with the audit strategy of DG AGRI and
the CAP single audit/control pyramid.
4
(b)
(i) The key and ancillary control lists were only agreed with MSs in the spring of 2015, hence not
available for CBs when guidelines were drafted. The new guidelines for FY2018 provide for the use
of the key and ancillary controls lists that have been updated for the period 2014-2020.
(ii) Taking into account the first year experience, the new guidelines for FY2018 address more
precisely these difficulties.
(iii) The new guidelines for FY2018 address this issue.
(iv) Requiring or allowing the CBs to go on-the-spot for transactions which have not been
controlled on-the-spot by the PAs would not be in line with the CAP single audit model and would
not bring comparable results vis-à-vis the PA's checks. Moreover, and especially for non-IACS
transactions, checking on-the-spot transactions previously not controlled on-the-spot by PAs would
generally not result in valuable additional evidence. The new guidelines for FY2018 provide
guidance to CBs on how to strike the right balance between costs and benefits in their work.
(v) The new guidelines for FY2018 address this issue.
(vi) The new guidelines for FY2018 address this issue. As provided for in them, the overall error to
be calculated by the CB for legality and regularity of the IACS transactions will be based on re-
verifications of on-the-spot checks, to be extrapolated to the remaining population.
The IRR represents the level of errors that were not detected by the PAs. Where the Commission is
satisfied that it can rely on the opinion of CBs, the overall rate of errors in the expenditure is
calculated as the sum of the error rate reported by the PA and the IRR estimated by the CB. This
logical approach of topping up the errors reported by PAs with estimates of non-detected errors has
been used in previous years by the Commission in DG AGRI Annual Activity Report.
INTRODUCTION
6. The Commission would like to precise that the minimum rate of on-the-spot checks can be
reduced down to 3% or even down to 1% under specific conditions, notably that the CB has
certified an error rate below 2%. In particular, a reduction to 1% may apply to basic payment
scheme, single area payment scheme, redistributive payment and small farmers' scheme.
12. The legal framework requires the CBs to issue an opinion drawn up in accordance with
internationally accepted audit standards. Moreover, the legislator empowers the Commission to
adopt implementing acts laying down rules on the tasks of the CBs, including the audit principles
and the audit methods (for instance the use of a single integrated sample). To that end, Article 6(4)
of Regulation (EU) No 908/2014 provides that the Commission shall establish guidelines to provide
further clarification and guidance in respect of the certification audit to be performed, and the
determination of the reasonable level of audit assurance to be achieved from audit testing.
OBSERVATIONS
22. The Commission considers that the CAP assurance model complies with the description of the
single audit as laid down by the Court and is a very good example of an effective control framework
in which each layer builds on the assurance provided by others. The model is based on solid
foundations that include accredited PAs managing the financial relationship with both beneficiaries
and the Commission, as well as comprehensive, detailed and precise rules on the management and
controls systems to be set-up (e.g. Land Parcel Identification System – LPIS). The effectiveness of
the model is evidenced by the relatively low error rates for EAGF and the continuous decrease in
the last 4 years in EAFRD error rates.
5
23. Common Commission's reply to paragraphs 23 and 24: For the CAP expenditure, the
Commission has built an assurance model that integrates the different layers of controls. The
control pyramid below shows how assurance is built upon the work carried out by the previous
layer, starting with the PAs' procedures and embedded controls in the internal control system, as
well as assurance from the internal audit service of the PA. In the Commission's model, the CBs'
work is based on the PA internal control system testing and reliance, as well as on their own
substantive testing.
The CAP single audit model
The Commission uses the results of its own audits on PAs to adjust (top-up) the error rate reported
by PA with an amount representing the level of errors not detected by PAs during their primary
controls before payments. The adjusted error rates are reported, at the level of each paying agency,
in the Annual Activity Report of DG AGRI and, together with tailored reservations and action
plans, are the basis for the declaration of assurance of the Director General of DG AGRI.
25. DG AGRI Audit Strategy 2014-2020 identified the new role of the CBs in giving an opinion on
legality and regularity as an opportunity to achieve more synergies with that of DG AGRI’s own
assurance building work reported in the Annual Activity Reports. This implies a gradual refocusing
of DG AGRI audits towards the CBs instead of the PAs in the mid and long term. However, the
audit work of the CBs will have to be checked before it will be possible to rely on the assurance
reported by them as regards legality and regularity. The objective is to place more and more
reliance on the assurance that can be obtained from the opinion of the CBs on legality and
regularity, as soon as it is verified that their work is to standard and can be relied upon.
27. The model described by the Court in Figure 2 is only one of the many possible models. The
Commission has chosen a more integrated approach taking the form of a pyramid of controls where
all layers of control give valuable evidence on both amounts and source of errors, thus maximising
their cost effectiveness. See Commission's reply to paragraphs 23 and 24.
28. The Commission welcomes the Court's suggestions on how the work of the CBs could improve
its assurance on the legality and regularity of EU expenditure and stresses that those have been
already implemented to the extent possible, i.e. when the work of CBs is considered reliable. For
instance:
First indent:
the Commission already uses the CBs' work for its adjustments (top-ups) to the error rates reported
in DG AGRI Annual Activity Report and which are close to those reported in the annual report of
the Court.
Second indent:
conformity clearance procedures are systematically launched as a result of the CBs' work in case the
errors exceed 50 000 EUR or 2% of the expenditure.
6
Third indent:
extrapolated error as calculated by the CB on the basis of their audit tests and statistical samples are,
where possible, used by the Commission for determining the amounts of financial corrections.
Fourth indent:
observations from the CBs can trigger requests to the competent authorities to review the
accreditation status and supervise the implementation of remedial actions.
30. When DG AGRI auditors identify deficiencies in the control system of PAs, they estimate the
amount at risk that was not detected by the PA and add it as a top-up to the error rate reported by
the PA. The result of the work of CBs on annual accounts was also used for top-ups in previous
Annual Activity Reports. The adjusted error rates are the basis for the declaration of assurance by
the Director General of DG AGRI.
31. DG AGRI could only make a limited use of the CBs' opinions for 2015 because in many cases
the work done was not yet reliable.
32. The Commission confirms the reasons why it could only make a very limited use of the
opinions in the first year of implementation of the new opinions on legality and regularity and refers
to DG AGRI Annual Activity Report 2015 for more information on the underlying reasons such as:
late appointment, late timing of the on-the-spot re-verifications, lack of experience, etc.
The Commission considers that, whereas the guidelines for FY2015 can be improved, where they
are correctly applied they already in their current form enable the CBs to give an opinion on legality
and regularity in compliance with applicable rules and standards.
33. The CAP assurance model (cf figure of pyramid of controls in the reply to paragraph 23 and 24)
is based on the control results of the PAs. The audit strategy of DG AGRI clearly sets out that the
ultimate objective for DG AGRI would be to get the necessary assurance from the audit work of the
CBs. The Commission considers that this assurance can be obtained where the CBs work is
considered reliable and their opinions are built upon the work carried out by the previous layer (i.e.
the PA's procedures and controls). Where the CBs’ work cannot be relied on, DG AGRI continues
to obtain independent assurance on legality and regularity of expenditure from its own audits.
34. In a single audit model the CB auditors should report on the errors (not) detected on the sampled
PA checks. PAs receive on an annual basis millions of aid applications, 100% of which are checked
administratively. In addition, PAs conduct on-the-spot controls before authorising the payment. For
most schemes, the on-the-spot check control rate is 5%, however, for some is 30% or even 100%. If
after its checks of the primary controls, a CB considers that the PA's procedures do not give
assurance on legality and regularity then it will give a qualified opinion, also affecting the error rate
provided by the PA.
35. The Commission's guidelines have been developed with the view to integrate the work of the
CBs in the CAP pyramid of controls, following a single audit approach, in accordance with EU
rules and ensuring that Member States could build on and further develop the methods and systems
already in place.
First indent:
See Commission's replies to paragraphs 48 to 58 and Recommendation 6.
Second indent:
See Commission's replies to paragraphs 62 to 67 and Recommendation 5.
Third indent:
See Commission's replies to paragraphs 68 to 71 and Recommendation 5.
7
Fourth indent:
See Commission's replies to paragraphs 72 to 78 and Recommendation 6.
Fifth indent:
See Commission's replies to paragraphs 79 to 85 and to Recommendation 2.
36. The Commission considers that, in accordance with internationally accepted audit standards, the
structure of the opinions shall reflect the different management and control systems.
There are two distinct agricultural funds: EAGF and EAFRD on which the CBs need to issue audit
opinions. In turn, the two Funds are divided into IACS1 and Non-IACS expenditure, depending on
the management and administration of the schemes/measures under the Fund. The IACS
schemes/measures are processed through a system with automated cross-checks embedded in it and
the system follows the whole cycle of the claim until the payment. In contrast, the Non-IACS
measures are not managed in one system. Furthermore, the timing of the controls in non IACS and
IACS is different. These inherent differences in the design of the schemes/measures prevents the
CBs from testing all measures in one Fund altogether. In order to comply with internationally
accepted audit standards (e.g. ISA315- Identifying and assessing the risks of material misstatement
through understanding the entity and its environment), the CB needs to first understand the
environment, identify risks and tailor appropriate audit procedures to cover those risks. In order to
be able to deliver a solid opinion on legality and regularity at Fund level, the CBs should assess the
risks separately for IACS and Non-IACS and design different audit procedures to cover those risks.
Thus, the guidelines on the Audit Strategy that the CBs are encouraged to use are in accordance
with internationally accepted audit standards. In addition, this approach serves the very purpose of
the opinion on legality and regularity, i.e. to ultimately and where necessary take appropriate
management decisions.
37. Please see Commission's replies paragraphs 37 to 84.
Common Commission's reply to paragraphs 41 to 44:
To ensure some continuity with the previous guidelines and enable CBs to build on the experience
from work done over many years for the annual financial clearance of accounts guidelines for
FY2015 did not substantially modify the way to use the accreditation matrix to form an opinion on
whether the PAs’ internal control systems are functioning properly. The accreditation matrix is
based on the accreditation criteria as laid down in Annex I of Regulation (EU) No 907/2014. Those
accreditation criteria are inspired from the COSO model which is recognized as a leading
framework for designing, implementing, and conducting internal control and assessing the
effectiveness of internal control.
The 2015 Guidelines were based on an integrated approach and sampling, in line with Article
9(2)(b) of Regulation (EU) No 1306/2013 and under which the testing, error evaluation and
reporting of errors under legality and regularity is combined with those for the annual accounts. In
this context, the usage of the whole accreditation matrix was deemed appropriate.
Notwithstanding the above, the Commission agrees with the Court that whilst the accreditation
criteria provide a solid basis for assessing the overall control framework of the PAs (what the audit
standards refer to as entity-level controls), the internal control system assessment for legality and
regularity on one hand, and, for the annual accounts2 on the other, needs to be separated.
1 Integrated Administration and Control System.
2 Execution of payments, accounting, advances and securities, debt management.
8
The approach to clearly separate the processes/procedures linked to legality and regularity from the
ones linked to the annual accounts is followed in the new guidelines for FY2018, which were
finalised in January 2017. This involves also amendment in the way the internal control system is
assessed for the purposes of the annual accounts and for the purposes of legality and regularity.
45. The Commission agrees with the Court. Similar observations were made by the Commission in
the letters sent to several MSs as a result of the 2015 financial clearance exercise.
Moreover, the Commission recommends to the CBs to look at action plans, which aim at addressing
known weaknesses. This link was discussed at the Expert Group Meeting of June 2016.
Nonetheless, this point is further emphasised in the new guidelines for FY2018.
In any event this must also be seen in the light of this being the first year that the CBs carried out
the new work on legality and regularity.
Box 1 – Situations where the accreditation matrix did not produce reliable results for legality
and regularity purposes
Please see Commission's reply to paragraph 45.
The weakness such as the one identified in Bavaria is provided for in the 2015 guidelines – section
6.5. In case the CB finds that the results of the substantive testing do not match with the previous
assessment of the Internal Control System (ICS), then the sample should be enlarged to
accommodate the real situation of the internal control system.
46. Comprehensive and detailed lists of key and ancillary controls have been elaborated by the
Commission in close cooperation with the MSs, for each of the type of measures of the CAP 2014-
2020. They are to be used at each control level.
47. The accreditation criteria should cover the overall set-up and functioning of the PA (i.e. the
entity-level controls), among other things, the appropriateness of processes and procedures put in
place.
When assessing the administrative and on-the-spot (OTS) process (the authorisation of claims) for a
given population, the CB's assessment should be based on the key and ancillary controls. The CBs
are also using the key and ancillary controls when they assess the legality and regularity of the
expenditure (under their compliance and substantive testing).
Nonetheless, the fact the CBs need to assess the process and procedures of authorisation of claims
(including administrative and on-the-spot controls) on the basis of the Key and ancillary controls is
clearly provided for in the new guidelines.
49. The split of the work for substantive testing into two samples resulted from the need to ensure
some continuity for Member States between the guidelines to CBs that were applicable before 2015
and those applicable from 2015. It also took into account the possibility established by Article
9(2)(b) of Regulation (EU) No 1306/2013 to use where appropriate a single integrated sample.
However, the experience of the first year of implementation showed that it was necessary to review
the approach and clearly distinguish the audit work and opinion on legality and regularity from the
opinion on the accounts, whilst keeping the possibility of dual purposes samples where appropriate
and with a view to avoid unnecessary administrative burden.
51. For sample 1 that is extracted from transactions controlled on-the-spot by the PAs, guidelines
require that CBs re-perform all controls done by the PAs, i.e. that CBs systematically re-perform the
on-the-spot controls. This approach allows CBs firstly to give an opinion on the control system and
secondly to calculate an incompliance rate (IRR) reflecting the difference they found between the
results of their re-performance and those of the primary controls by the PAs.
9
52. On the basis of the internal control system assessment, CBs should determine the level of
substantive testing (both sample 1 and sample 2). Thus, it is a CB's prerogative to increase the level
of substantive testing in order to cover the risks. In the table 3 presented by the Court, it is shown
that some CBs decided to increase the samples of substantive testing above the minimum 30
transactions in order to address the high risks in the given population.
55. The CBs are required to make their sample selection on claims for the IACS populations due to
the changing nature of the agricultural reality. Thus, it is crucial that the CB goes on the spot shortly
after the PA in order to be able to re-verify all eligibility elements. If re-verification is performed
late, there is a huge risk that not all eligibility elements can be verified or the reality has changed
completely, making re-verification ineffective and the opinion unreliable.
56. Due to the agricultural reality, there is indeed a risk related to the timing of the primary on-the-
spot checks and their re-verifications.
The guidelines address that risk and the CBs were repeatedly invited by the Commission to
adequately plan their re-verifications taking into account the time constraints imposed by the nature
of the re-verifications.
As reported in DG AGRI Annual Activity Report 2015, inadequate timing of the re-verifications
was unfortunately one of the main reasons why the Commission was not able to rely on many of the
CBs' opinions on 2015 expenditure.
Box 3 – Delays in the CBs’ re-performance of IACS Sample 1 on-the-spot checks for the
2014 claim year (2015 financial year)
Delays in appointing the CBs and in re-verification of the controls in claim year 2014 (financial
year 2015) were the main reasons why the Commission could not rely on CBs' opinions on legality
and regularity for its Annual Activity Report 2015.
Already during its 17 audits done in 2015, the Commission has observed in numerous cases
significant delays in the re-performance of controls. The Commission has issued a Note to the MSs
on 09/11/2015 in which it explains that the on-the-spot re-verifications should start in the summer
of n-1 for financial year n.
57. Under the FY2015 guidelines, the CB's re-verification sample is to be drawn from the list of
transactions selected for random on-the-spot check by the PA. Thus, the CB's selection and actual
re-verification should normally be done before the PA starts processing payments. As a result, it is
very unlikely that the CB will not select an item because the PA is delaying the payment.
Box 4 – Exclusion of transactions from IACS Sample 1 because the PA did not execute
payment until after the CB had completed its audit work
The work of the CBs on legality and regularity is threefold:
- opinion on the effectiveness of the internal control system;
- opinion on the legality and regularity of expenditure;
- opinion on the assertions made in the Management declaration (i.e. the control results reported by
the PA).
According to the guidelines, the CB should have still used these five transactions when assessing
the assertions reported in the Management Declaration, as well as on the effectiveness of the
internal control system. However, Article 9 of Regulation 1306/2013 requires an opinion on the
legality and regularity of the expenditure. Thus, the CB cannot issue an opinion on non-executed
payments. In this case, an additional item should have been tested to comply with the originally
determined sample size for the financial year in question. In order to anticipate such events, the CBs
10
are advised to select some additional items in advance, above the minimum 30 foreseen in the
guidelines.
The CB should consider the result of these tests for the opinion on the legality and regularity of
expenditure regarding the financial year in which this transaction will be paid. Those tests should be
taken into account irrespective of any correction the PA could do before payment.
58. Article 7 of Regulation (EU) No 908/2014 allows the CBs to accompany the second-level on-
the-spot checks. Accompaniment of the first-level checks is limited to the check that could not be
re-performed at a later stage (such as EAGF Non-IACS measures which involve operations linked
to distillation, storage or withdrawal operations). The idea behind the re-performance is that the PA
should do the check, so that the CB re-verifies it afterwards.
Notwithstanding the above, the guidelines request CBs to perform their tasks following professional
audit standards and not compromising the results of the substantive testing. Thus, the CB should
never share its selection before the PA conducts its checks. The shortcoming described by the Court
in Box 5 confirms that CBs should follow the guidelines to avoid inadequate audit practices.
Box 5 – The re-performance exercise was compromised in Italy because the CB had given the
PA advance notice of which beneficiaries would be subject to re-performance
CBs shall put in place safeguards for not disclosing their samples to the PAs before the latter have
concluded their own checks and advised the CBs accordingly.
61. There is a mismatch between the controls performed by the PA in a given year and the
payments for the financial year for the Non-IACS populations. Nonetheless, since the majority of
the Non-IACS measures are not time-constrained, the CB could test additional transactions in case
some of the already tested ones were not paid during the year. Unpaid transactions could still be
used to confirm or not the Management declaration and the control results.
62. In the CAP framework, where a legally binding and effective management and control system is
to be put in place by the PAs, auditors can in many cases, once they have sufficient assurance that
the PA's management and control system is designed and implemented in accordance with EU law,
gather sufficient and appropriate evidence by re-performing the PA controls.
Common Commission's reply to paragraphs 64 and 65:
For sample 1 that is used for establishing the IRR, guidelines require CBs to re-verify both
administrative and on-the-spot primary controls. For sample 2 that is used for establishing the EER,
the Commission's guidelines do not require on-the-spot checks. The proportion between the two
samples is to be defined by the CBs following their professional judgement.
66. The Commission considers that re-performance of administrative checks is sufficient and
appropriate audit evidence. It does include inquiry to the PA's personnel as to the checks performed,
checks of the aid application and supporting documents, as well as checks in registers and LPIS for
IACS schemes. For non-IACS expenditure, administrative checks and their re-verification also
include in-situ visits (inspections) to, for instance, check the existence of physical assets built or
purchased through investment projects.
It is up to the CB's professional judgement to design the audit procedures in order to gather
sufficient and appropriate audit evidence. Section 6.1 of Guidelines 2 for FY2015 says: "The test of
detail on operational transactions – includes, among others…These cover inspection and enquiry,
not only documentary review".
67. For testing the legality and regularity of the IACS expenditure, CBs have to select their sample
from the population that was controlled on-the-spot by the PAs and to re-perform all the primary
controls, both administrative and on-the-spot.
11
For non-IACS expenditure, it should be noted that most of the errors (ineligibility of
beneficiaries/activities/projects/expenditure and non-compliance with public procurement rules) are
found at the level of administrative controls and that around two thirds of the errors came from
insufficient use of available information or errors by the national authorities. It should also be noted
that all administrative controls (to be re-performed by the CBs) on investment operations must
include an in-situ visit to verify the realisation of the investment, except if the paying agency
justifies that the operation is small, low risk or is included in the on-the-spot sample.
Furthermore, taking into account the need to provide guidance to CBs on how to strike the right
balance between costs and benefits, the Commission's guidelines for FY2018 request CBs to have
statistically representative samples for the on-the-spot re-verifications for IACS. For non-IACS
expenditure a minimum of 30 on-the spot re-verifications are requested. The results of the re-
verifications will be extrapolated to the entire population.
69. The Commission considers that under the CAP integrated management and control system,
assessing the checks of the PA as regards legality and regularity, and afterwards, confirming from a
sample of transactions, that the PA has effectively implemented the procedures for primary
controls, is sufficient to provide assurance on the legality and regularity of the expenditure.
71. The Commission considers that in the CAP framework, checking the control systems and re-
performing the primary controls of the PAs generally allows obtaining sufficient and appropriate
evidence and forming a reasonable assurance opinion on the CAP expenditure.
In addition, the Commission guidelines do not prevent CBs from doing any audit work that they
would themselves consider appropriate. It is CBs' role to determine if re-performance is enough.
Inquiry and inspection may also be used in the re-verification of administrative checks. Cf
Commission's reply to paragraph 66.
74. Due to the integrated approach used in the current guidelines, CBs were requested to calculate
two rates of error that deal with legality and regularity. Nonetheless, this approach is revised from
FY2018.
In the new guidelines that are applicable from FY2018 and were presented in January 2017, the
conformity and financial clearance procedures are clearly separated. To that end, CBs are to
calculate two separate rates of error: one error rate (ERR)3 – including only errors from the annual
accounts testing, and one incompliance rate (IRR)4 including all legality and regularity errors. The
Commission already foresees in its new guidelines for FY2018 that the IRR will be used both for
the audit opinion on legality and regularity and to confirm (or not) the assertions in the Management
declaration.
Common Commission's reply to paragraphs 75 and 76:
In cases where there is sufficient evidence that CBs have worked in accordance with the guidelines,
the Commission uses the IRR estimated by the CBs in its assessment of the top-up to the error rate
calculated on the basis of the control statistics reported by MSs at the level of each PA. In addition,
the ERR was taken into account for assessing the level of the top ups.
77. With the new guidelines from FY2018, the ERR includes only errors from the annual accounts
testing and will be used only to assess the reliability of the PAs' accounts.
3 ERR – error rate used to estimate the misstatements in the audited sample and population with a financial impact on
the completeness, accuracy and veracity of the accounts. 4 IRR – incompliance rate estimating the potential financial impact due to deficiencies in the eligibility controls.
12
78. The Commission draws the attention to the integrated sampling foreseen by the legislation and
confirms that it separates the financial and conformity clearance procedures. The new guidelines for
FY2018 clarify it further, with an error rate for legality and regularity and another one for reliability
of the annual accounts.
Common Commission's reply to paragraphs 79 to 84:
As reported in DG AGRI's Annual Activity Report, Annex IV, materiality criteria, the Commission
uses control statistics reported by MSs to calculate a "residual reported error rate" by extrapolating
to the entire population the errors found by PAs in the population randomly selected for on-the-spot
controls. This residual reported error rate is incomplete because it does not take into account the
risk that deficiencies in the controls done by PAs might result in errors not being detecting and
reported.
Therefore, in a second step, DG AGRI collects all available information that would reveal
deficiencies in the first level controls, including the results of its own audits, of ECA's audits or any
other reliable source of evidence. DG AGRI auditors analyse this information, in the light of the
Commission's guidelines on the calculation of the financial corrections (Document C(2015) 3675
final) and using their professional judgement, and adjust the residual reported error rate with a top-
up that estimates the rate of errors not detected by PAs. The Commission uses the adjusted error
rate for estimating the amounts at risk and analysing whether reservations are necessary.
From 2015, the CBs' opinions on legality and regularity of the expenditure constitute potentially
another source of audit evidence for the Commission. In particular, the IRR may be used by the
Commission as a reliable estimate of the top-up for adjusting the error rate for the expenditure
under each PA. In addition the ERR was taken into account for assessing the top ups. However, in
2015 the opinions were often not yet sufficiently reliable for the Commission to use them to a large
extent.
The Commission considers that this use of the IRR to reinforce its assurance on the legality and
regularity of the CAP expenditure is appropriate and that, when established in accordance with its
guidelines, the use of the IRR to adjust the error rate does not understate the total error.
85. The Commission is aware of these potential risks that the opinion of the CBs on legality and
regularity would not be reliable and considers that they are covered in its guidelines. The
Commission will continue to supervise closely that CBs follow the guidelines as well as the audit
standards, including the application of their professional judgement, and conclude on legality and
regularity with the necessary qualification.
CONCLUSIONS AND RECOMMENDATIONS
Common Commission's reply to paragraphs 89 and 90:
The pyramid of controls of the CAP expenditures, which already corresponds to the single audit
approach described by the Court in paragraph 22, has been complemented from FY2015 with the
integration of the CBs between the PAs and the Commission layers, reinforcing the CAP assurance
model. In accordance with Article 6(4) of Regulation (EU) No 908/2014 the Commission has
established guidelines to provide further clarification and guidance in respect of the certification
audit to be performed, and the determination of the reasonable level of audit assurance to be
achieved from audit testing. The Commission considers that, whereas the guidelines for FY2015
can be improved, where they are correctly applied they already in their current form enable the CBs
to give an opinion on legality and regularity in compliance with applicable rules and standards.
Building on the experience from the first year of implementation, the Commission has elaborated
new improved guidelines to be applied for FY2018 onwards.
13
92. The Commission is satisfied with the CAP assurance model that keeps the error rate under
control at relatively low levels. The overall error rate for Natural resources (where the CAP
represents 98 % of the expenditure) established by the Court for financial year 2015 was 2.9 %. For
EAGF, where payments are granted every year to close to 8 million beneficiaries, DG AGRI
estimates that the error rate has been below materiality in the last two years, while the Court's most
likely error (MLE) was stable at 2,2 %. For EAFRD, the level of errors established by the
Commission has persistently decreased in the last 4 years.
Such results would not be achievable without the solid management and control system of the CAP,
notably IACS and the pyramid of controls where assurance is built upon the work carried out by the
previous layers and where the work of the CBs starts from the PAs' control results. In this respect,
CBs are requested to confirm the results of the controls performed by PAs and, if not confirmed, to
establish an incompliance rate and to give a qualified opinion. However, DG AGRI could only
make a limited use of the CBs' opinions for 2015 because in many cases the work done was not yet
reliable.
Where the CBs’ work cannot be relied on, DG AGRI continues to obtain independent assurance on
legality and regularity of expenditure from its own audits and where necessary used them as a basis
for top ups to the error rates reported by the PAs. The resulting adjusted error rates remained
relatively low (cf table 1 and above).
Where the Commission can rely on the CBs' work, the incompliance established by CBs can be
added to the error rate reported by PAs to calculate a reliable error rate that reflects the level of
assurance on the legality and regularity of CAP expenditure.
See also Commission's reply to Recommendation 1.
Recommendation 1 - Assurance from the CBs’ work on legality and regularity
The Commission partially accepts the recommendation, as follows:
The Commission does not accept the recommendation that using the opinion of CBs for
assurance would be possible only once the work is carried out as defined in recommendations 2 to
7, and not where the work is done as defined in the guidelines applicable from the financial year
2015. The Commission considers that, whereas the guidelines for FY2015 can be improved, where
they are correctly applied they already in their current form enable the CBs to give an opinion on
legality and regularity in compliance with applicable rules and standards.
The Commission accepts the recommendation that where the audit work of the CBs is done in
accordance with the applicable regulations and guidelines their opinion are an important and
valuable additional building block and should be the key element for the Commission's assurance.
This is in line with the audit strategy of DG AGRI and the CAP single audit/control pyramid below:
The CAP single audit model
14
93. The accreditation criteria should cover the overall set-up and functioning of the PA (i.e. the
entity-level controls), among other things, the appropriateness of processes and procedures put in
place. However, when assessing the administrative and on-the-spot (OTS) process (the
authorisation of claims) for a given population, the CB's assessment should be based on the key and
ancillary controls. Thus, the key and ancillary controls that define the controls for the particular
population (or scheme/measure) should be used on top of the overall accreditation criteria. The CBs
are also using the key and ancillary controls when they assess the legality and regularity of the
expenditure (under their compliance and substantive testing). (Cf. Commission's reply to paragraphs
41-44).
Nonetheless, the fact the CBs need to assess the process of authorisation of claims (including
administrative and on-the-spot controls) on the basis of the Key and ancillary controls is made clear
in the new guidelines for FY2018.
94. The Commission acknowledges that whilst the accreditation criteria provide a solid a basis for
assessing the overall control framework of the PA (what the audit standards refer to as entity-level
controls), the internal control system assessment for legality and regularity and for the annual
accounts 5 need to be separated. (Cf. Commission's reply to paragraphs 41-44).
The approach to clearly separate the processes/procedures linked to legality and regularity from the
ones linked to the annual accounts is adopted in the new guidelines for FY2018, which were
finalised in January 2017.
Recommendation 2 – Risk assessment focused on key and ancillary controls
The Commission accepts the recommendation and considers it already implemented in the new
guidelines for FY2018. The CBs are advised to use the key and ancillary controls when testing the
procedures for authorisation of claims.
95. The Commission stresses that sample 1 is used not only to test the functioning of the internal
control system and the reliability of the control statistics, but also to conclude on the legality and
regularity of expenditure at beneficiary level. Nonetheless, the new guidelines for FY2018 requires
a separation of audit objectives where the sample for legality and regularity will be used for the
following purposes:
- to conclude on the on the legality and regularity of expenditure;
- to conclude on the control results reported in the Management Declaration and the control
statistics;
- for the internal control system testing if dual-purpose testing is done by the CB;
5 Execution of payments, accounting, advances and securities; debt management.
15
- to certify an error rate for the purposes of the on-the-spot control rate reduction in accordance with
Article 41 of Regulation (EU) No 908/2014.
For IACS expenditure, the applicable legislation and the Commission guidelines to PAs include
methods aimed at assuring the representativeness of the sample with regards to the total population.
In the first step of their sampling operations, the CBs have to check that the PAs' samples are legal
and regular, i.e. that they have been selected in accordance with EU rules and guidelines to PAs,
including representativeness. In turn, the CBs' samples need to be representative for the population
selected by the PAs.
96. The issues related to the timing of the on-the-spot checks and the re-verifications for IACS
transactions are inherent to the changing agricultural reality in the fields.
Thus, it is crucial that the PAs do the primary control at the right time, i.e. before payments, and
that CBs go in the field shortly after the PAs in order to be able to re-verify all eligibility elements.
If the re-verification is performed late, there is a risk that not all eligibility elements will be verified
in the field or the reality has changed completely, so a re-verification will not be possible.
Recommendation 3 – Safeguards for IACS sampling based on PAs’ on-the-spot checks
The Commission accepts the recommendation and considers that it is already implemented in the
new guidelines for FY2018. The new guidelines are reinforced as regards the following aspects:
First indent:
the CB should check that the PA sample is selected in accordance with relevant legislation, rules
and guidelines including representativeness and then, the CB should check the representativeness of
its own sample;
Second indent:
crucial importance of the timing for the IACS re-verifications: a recommended timing of the audit
activities is included in the guidelines;
Third indent:
as part of the audit procedures, the CB should check the completeness and assurance of the control
results reported in the control statistics;
97. The Commission agrees that there is a mismatch between the controls performed by the PA in a
given year and the payments for the financial year for the Non-IACS populations. Nonetheless,
since the majority of the Non-IACS measures are not time-constrained, the CB could test additional
transactions in case some of the already tested ones were not paid during the year. The unpaid
transactions could still be used to confirm or not the Management declaration and the control
results, as well as the effectiveness of the internal control system.
Recommendation 4 – Non-IACS sampling based on payments
The Commission accepts the recommendation and considers that it is already implemented in the
new guidelines for FY2018. In case of time constraints, e.g. for measures entailing many payments
close to the end of the year, the CB is recommended to use estimated payments.
98. The Commission considers that in the CAP framework and notably when taking into account
the IACS contribution to preventing and reducing the level of errors, re-performance of
administrative checks is appropriate audit evidence. It does include inquiry to the PA's personnel as
to the checks performed, checks of the aid application and supporting documents, as well as checks
in registers and LPIS for IACS schemes. For non-IACS expenditure, administrative checks and
16
their re-verification by CBs also include in-situ visits (inspections) to, for instance, check the
existence of physical assets built or purchased through investment projects.
The Commission also considers that re-performing the checks done by PAs is generally sufficient
for CBs to assess the legality and regularity of the expenditure. To check on-the-spot payments not
controlled on-the-spot by PAs would generally not result in valuable additional evidence and will
generate incomparable results vis-à-vis the PA' controls.
99. The guidelines should be seen as the minimum that is necessary to satisfy the Commission and
cannot limit CBs from doing more if appropriate.
The Commission considers that in the context of the CAP expenditure and its management and
control system, re-performance is the most efficient way to reinforce assurance and to form a
reasonable assurance audit opinion.
As explained in the Commission's replies to paragraphs 60-70, the Commission's guidelines
recommend to the CB to perform the same checks (on-the-spot check and administrative checks or
only administrative checks) as the PA so that comparable results are obtained. Moreover, the
guidelines do not limit the CBs to do re-performance but re-performance in combination with other
techniques so that sufficient and appropriate audit evidence is collected. Since the CBs re-verify
also the administrative checks, inspection of documents and registers, as well as inquiry, are also
common techniques used by them.
In the new guidelines for FY2018, the Commission recommends the CBs to have two samples
dedicated to the separate audit objectives (audit of the accounts and audit of legality and regularity).
The Commission has already identified and took on board an approach that allows for an efficient
and effective audit such as: dual-purpose testing between audit objectives and between substantive
and compliance testing6, etc. The audit technique is not limited to re-performance but to the scope
of the PA's checks. Thus, if a PA only checked a transaction administratively, the CB should do the
same – re-verify the administrative checks through re-performance, inspection, inquiry, analytical
checks, etc.
Taking into account the need to provide guidance to CBs on how to strike the right balance between
costs and benefits, the Commission's guidelines for FY2018 request CBs to have statistically
representative samples for the on-the-spot re-verifications for IACS. For non-IACS expenditure a
minimum of 30 on-the spot re-verifications is requested. The results of the re-verification will be
extrapolated to the entire population.
Recommendation 5 – Substantive testing carried out on-the-spot
The Commission partially accepts the recommendation as follows:
1) The Commission does not accept the recommendation where it would entail that the CBs go
on-the-spot for transactions which have not been controlled on –the-spot by the PA. This would not
be in line with the Commission’s CAP assurance model and would not bring comparable results
vis-à-vis the PA's checks. Moreover, and especially for non-IACS transactions, checking on-the-
spot transactions previously not controlled on-the-spot by PAs would generally not result in
valuable additional evidence.
Taking into account the need to provide guidance to CBs on how to strike the right balance between
costs and benefits, the Commission's guidelines for FY2018 request CBs to have statistically
representative samples for the on-the-spot re-verifications for IACS. For non-IACS expenditure a
minimum of 30 on-the spot re-verifications are requested. The results of the re-verification will be
6 The possibility to use one transaction for various audit tests.
17
extrapolated to the entire population. The Commission considers that in the CAP framework this
balanced approach would be sufficient for the CBs to deliver a reliable opinion on the legality and
regularity of expenditure.
2) The Commission accepts the recommendation as regards allowing the CBs to carry out all the
audit steps and procedures that they themselves consider appropriate and considers that it is already
implemented.
100. Due to the integrated approach used in the current guidelines, CBs were requested to calculate
two rates of error that cover legality and regularity. Nonetheless, this approach is revised in the
guidelines for FY2018 (cf. Commission's reply to Recommendation 6 below).
101. To ensure continuity, the guidelines for 2015 – first year for the enhanced work of the CBs -
were elaborated on the basis of guidelines used for previous years' financial clearance. Furthermore,
the 2015 Guidelines were based on an integrated approach and sampling, in line with Article 9(2)(b)
of Regulation (EU) No 1306/2013 and under which the testing, error evaluation and reporting of
errors under legality and regularity is combined with those for the annual accounts. It follows that
under the integrated approach, the ERR is used for assessing legality and regularity of the
expenditure, as well as the annual accounts.
Recommendation 6 – A single error rate in relation to legality and regularity
The Commission accepts the recommendations and considers that it is already implemented.
The new guidelines for FY2018 foresee to clearly separate the conformity and financial clearance
procedures. To that end, CBs are to calculate two separate rates of error: an error rate (ERR)
including only errors from the annual accounts testing, and an incompliance rate (IRR) including all
legality and regularity errors not detected by the PA. The IRR will be used both for the audit
opinion on legality and regularity and to confirm (or not) the assertions in the Management
declaration. The IRR will be used to top up the control results reported by the PAs.
102. The incompliance rate (IRR) is designed to estimate by how much the PA's controls differ
from the CB's re-verification or in other words, how much the PA failed to detect. Thus, it is
reasonable to assume that if the PA finds an error, this error will be corrected before payment. The
CB's error should reflect only how much was not detected and as a result, not corrected before
payment. That is why in order to avoid double-counting of errors, the CB should not include the
beneficiary's claim, as the difference will include a part that was already corrected by the PA and
reported in the control statistics.
103. The CBs work is based on samples in accordance with internationally accepted audit standards.
In any event, all findings found during the on-the-spot re-verifications should be extrapolated to the
rest of the population for the calculation of the incompliance rate (whether or not they were
resolved before payment).
A similar approach is used by DG AGRI in the calculation of the residual error rate in the Annual
Activity Report.
Recommendation 7 – An error rate representative of the population covered by the audit
opinion
The Commission accepts the recommendation and considers that it is already implemented.
The new guidelines for FY2018 foresee that the incompliance rate for the IACS populations be
based on the PA's random on-the-spot check sample, however, the extrapolated result for the part of
the population is then extrapolated to the whole population in order to conclude on the legality and
regularity of all IACS expenditure.
For the Non-IACS populations, a slightly different approach is foreseen to take into account the
different design of the measures and the fact that the PA's random samples may not be
18
representative in all cases. Thus, the incompliance rate to be calculated by the CBs should be based
on a sample drawn from all expenditure which is then extrapolated.
104. The Commission considers that with the new guidelines for financial year 2018
recommendations 2, 3, 4, 6 and 7 are already implemented and that the part of recommendation 5
that is accepted is also already implemented. Since IACS related payments in financial year 2018
correspond to beneficiaries' claims introduced in Spring 2017 and to be checked during the same
crop period, the implementation should be realistic: implementing the new guidelines will require
preparatory work for the MSs, including public procurement procedures in cases where CBs are not
public bodies. Most MSs have already informed the Commission that they could apply the new
guidelines only from financial year 2019.
Event Date
Adoption of Audit Planning Memorandum (APM) / Start of audit 13.1.2016
Official sending of draft report to Commission (or other auditee) 6.2.2017
Adoption of the final report after the adversarial procedure 22.3.2017
Commission’s (or other auditee’s) official replies received in all languages
26.4.2017
This audit examined the role of the Certification Bodies which provide opinions on the legality and regularity of spending under the Common Agricultural Policy at Member State level. The Common Agricultural Policy accounts for almost 40 per cent of the EU Budget. We assessed whether a new framework set up in 2015 by the European Commission enables the Certification Bodies to form their opinions in line with EU regulations and international audit standards. Although the framework is a positive step towards a single audit model, we found that it is affected by significant weaknesses. We make a number of recommendations for improvement, to be included in new Commission guidelines due into force from 2018.
EUROPEAN COURT OF AUDITORS 12, rue Alcide De Gasperi 1615 Luxembourg LUXEMBOURG
Tel. +352 4398-1
Enquiries: eca.europa.eu/en/Pages/ContactForm.aspx Website: eca.europa.euTwitter: @EUAuditors
© European Union, 2017
Reproduction is authorised provided the source is acknowledged.
1977 - 2017